Frederick R. Singer

Incidence and Epidemiology

Studies of the incidence of Paget’s disease are inherently imprecise because many affected individuals are asymptomatic and have normal serum total alkaline phosphatase levels.¹ On the basis of autopsy studies^2,3 and review of radiographs, the prevalence of the disease is believed to be 3% or greater in individuals older than 40 years in countries where the disease is common.⁴

A striking feature of the epidemiology of Paget’s disease is the great variability in prevalence estimates in different regions of the world and even within a single country. A survey of hospital radiographs in patients older than 55 years in 31 British towns revealed a prevalence of Paget’s disease ranging from 2.3% in Aberdeen, Scotland, to 8.3% in Lancaster, England.⁵ A similar survey done throughout Europe found that only in France did the prevalence equal the lowest prevalence rates in Britain.⁶ Australia, New Zealand, and the United States⁷ have a relatively high prevalence, perhaps because of British migration. The disease appears to be rare in Asia. In Japan, the prevalence has been estimated to be 2.8 per million of the population.⁸ In most studies, the prevalence of Paget’s disease in men slightly exceeds that in women.^2–7 Several studies suggest that the prevalence rate of Paget’s disease is decreasing in Britain,⁹ New Zealand,¹⁰ and the United States.¹¹ Whether a decrease in prevalence has occurred or whether an ascertainment bias has been introduced by the use of automated serum chemistry panels about 30 years ago is unclear.

Much emphasis has been placed on the increase in prevalence of Paget’s disease with aging. It has been estimated that the prevalence is nearly 10% by the ninth decade.2–4 Conversely, the disease has rarely been reported in individuals younger than 20 years. Although Paget’s disease is most often recognized after age 50 years, it is probably misleading to conclude that the disease is rare in younger individuals. As is discussed later, the obvious manifestations, such as skeletal deformity, probably evolve over decades. Failure to diagnose the earlier phases of the disease is no doubt due to lack of symptoms and minimal use of the radiologic and biochemical tests that would lead to an early diagnosis in younger individuals.

Since 1883, it has been appreciated that Paget’s disease may affect more than one member of a family.¹² In large studies, a family history of the disease has been obtained in 12.3% of 788 cases¹³ (United States), 13.8% of 407 cases¹⁴ (Great Britain), and 22.8% of 658 cases¹⁵ (Australia). In the former two studies, a 7- to 10-fold increase in Paget’s disease was noted in relatives of patients in comparison with control groups. In a small study in Spain in which relatives of patients were screened with bone scans, 40% of patients had at least one first-degree relative with Paget’s disease.¹⁶ Examination of the overall pattern of apparent transmission suggests an autosomal dominant mode of inheritance.¹²

The search for potential environmental factors in the pathogenesis of Paget’s disease has led to consideration of whether past ownership of dogs might be a risk factor,¹⁷ but this has not been confirmed by subsequent studies. Occupational exposure to lead has been proposed as a possible factor in Paget’s disease.¹⁸ In one study, levels of cortical bone lead were higher than those in control subjects, but trabecular bone lead was lower.¹⁹ The relevance of these findings is unknown.

Pathophysiology

Consideration of the radiologic and pathologic evolution of the lesions of Paget’s disease strongly suggests that the primary disturbance is localized acceleration of osteoclastic bone resorption. At the interface of normal bone and an advancing lesion, numerous osteoclasts are found in Howship’s lacunae in cortical or trabecular bone.² Many of the osteoclasts are larger than normal and may have up to 100 nuclei in cross-section rather than the several found in normal osteoclasts.²⁰ Examination of the ultrastructure of osteoclasts in specimens of Paget’s disease reveals a striking and characteristic feature: the presence of microfilaments in the nucleus and, less frequently, in the cytoplasm.^21,22 These structures have not been observed in osteoclasts from normal subjects or in the bone of patients with primary or secondary hyperparathyroidism, osteoporosis, or osteomalacia. They also have not been observed in osteoblasts, osteocytes, or bone marrow cells in the lesions of Paget’s disease. The inclusions have been found in a small percentage of the multinucleated giant cells (osteoclasts) in giant cell tumors of bone and in the osteoclasts of some patients with osteopetrosis and pyknodysostosis.²³ The structure of the microfilaments most closely resembles the nucleocapsids of viruses of the Paramyxoviridae family, a group of RNA viruses known to cause common childhood infections. Evidence of Paramyxoviridae nucleocapsid proteins^24,25 and messenger RNA (mRNA)^26,27 has been found in pagetic lesions. The full-length sequence of the measles virus nucleocapsid gene has been sequenced from a pagetic lesion of one patient.²⁸ The relevance of these findings to the cause of Paget’s disease is discussed later.

As osteoclastic resorption progresses in the cortex, individual osteons widen and become confluent with adjacent osteons. The resorptive area thereby may extend to the endosteal and periosteal surfaces. In trabecular bone of the medullary cavities, the osteolytic process results in a marked reduction in bone volume. Associated with both early cortical and trabecular lesions is a remarkable proliferation of fibrous tissue that replaces normal fatty or hematopoietic bone marrow. The fibrous stroma is highly vascular, and although arteriovenous shunts were previously thought to be present, this feature has not been confirmed with radiolabeled albumin microspheres.²⁹

The earliest recognizable radiologic feature of Paget’s disease is a focal osteolytic lesion. The skull was first appreciated to be affected by circumscribed osteolytic lesions, and Schuller³⁰ applied the term osteoporosis circumscripta to this finding. One or more osteolytic foci may be present, most often in the frontal and occipital regions, and may be observed to coalesce slowly over a period of years (Fig. 16-1). Pure osteolytic lesions also may be detected in other regions of the skeleton, but less frequently. They are seldom observed in the vertebral column or pelvis. In the long bones, osteolytic lesions usually develop at either end of the bone, less often in the diaphysis.³¹ Occasionally, osteolytic foci can be observed simultaneously at both ends of a bone. The junction of normal bone and the osteolytic lesion shows a characteristic appearance that is nearly diagnostic of Paget’s disease. The edge of the lesion usually assumes the shape of a flame or an inverted V (Fig. 16-2). Serial radiologic follow-up of untreated lesions has documented an average rate of extension of about 1 cm annually.³¹ Bone biopsies taken during this earliest phase of Paget’s disease not unexpectedly reveal a marked increase in osteoclastic activity and thinning of the trabeculae.³² Other striking features include a fibrovascular marrow, numerous osteoblasts lining the trabeculae, and prominent woven bone. Thus, a discrepancy is found in the results of radiologic and pathologic examination of an osteolytic lesion. Although focal density is decreased on radiographs, histology demonstrates very active bone formation, but not sufficient to overcome the remarkable degree of osteoclastic bone resorption.

A more commonly observed stage of Paget’s disease is the mixed phase, in which osteoblastic (or osteosclerotic) features are intermixed with osteolytic features in an individual bone. This phase is best appreciated in long bones, in which one may observe the advancing osteolytic front adjacent to normal bone and, trailing this front, a heterogeneous region of osteosclerosis superimposed on the region that had previously been dominated by the osteolytic process (Fig. 16-3). Biopsies of the mixed phase reveal a characteristic abnormality of lamellar bone in both cortical and trabecular bone. The matrix is transformed into a bizarre “mosaic” pattern of irregularly juxtaposed pieces of lamellar bone separated by cement lines that have a scalloped outline. The irregularity probably reflects areas of previous osteoclastic resorption. The structure of the involved cortex is so disordered that complete osteons are rare, and the outer and inner circumferential lamellae and the interstitial lamella may be totally disrupted.

The same disordered matrix structure is seen in trabecular bone. Interspersed among the chaotic lamellae are patches of woven bone characterized by a random pattern of deposition of collagen fibers and a larger number of osteocytes per unit area of matrix. It has been suggested that the lacunae surrounding the osteocytes are larger than normal and that this finding represents osteocytic osteolysis.³³ However, it is more likely that the increased size of the periosteocytic lacunae is simply a characteristic of woven bone and not a second type of bone resorption in Paget’s disease. At the surfaces of bone formation, plump osteoblasts are found in great number adjacent to abundant osteoid. This type of bone is seldom found in adults except when associated with rapid remodeling of bone, such as occurs after a fracture or in response to tumor invasion of bone. Studies using quantitative histomorphometry of bone documented the marked degree of cellular activity underlying the dramatic changes in bone structure in Paget’s disease.³⁴ The total amount of osteoid and the percentage of the bone surface covered by osteoid may be increased fourfold to fivefold. The increase in osteoid is not associated with an increase in osteoid seam width because the rate of calcification also is increased, as established by double labeling of bone-forming surfaces with tetracycline. No dynamic means of defining the rate of bone resorption is available, but the extent of the total bone surface exhibiting evidence of bone resorption averages about sixfold that of normal individuals, and the number of osteoclasts may be increased as much as 10-fold. In the medullary cavities, the intense resorptive process may produce hemorrhagic cysts with encircling fibrous marrow containing macrophages filled with hemosiderin. These cysts are believed to result from rupture of multiple dilated vessels and ensuing microinfarctions.³⁵ In the mixed phase of Paget’s disease, not only does patchy sclerosis of bone become apparent on radiography, but a bone also may be enlarged. If the osteolytic process extends to the subperiosteal layer, bone formation may be stimulated to such an extent that the thickness and circumference of the bone are increased as a result of periosteal new bone formation. When the skull is affected, this process can produce as much as a fourfold thickening of the calvarium, as was reported by Paget.³⁶ A patchy form of sclerosis is often of a “cotton-wool” character (Fig. 16-4). The skull may be so severely affected that platybasia, or basilar impression, may be a complication. Long bones may be shortened, and typically lateral bowing of the femur or anterior bowing of the tibia or both may develop. Later in the course of the disease, the tibia may exhibit severe lateral bowing. The pathogenesis of the slowly progressive deformity is not known but must be related to the state of abnormal remodeling of the bone. Frequently, fissure fractures are associated with the bowing deformity. These fractures are linear transverse radiolucencies that usually are present in the cortex of the convex aspect of the deformed bone. They often are multiple and may remain stable in appearance for years. They can be found in either osteopenic or osteosclerotic cortices and may be present even in the absence of deformity. Histologic examination of these lesions suggests that they are incomplete fractures.² Only a small percentage of these lesions progress to a complete transverse fracture, which has been seen more often in patients with a sclerotic cortex.

Even after osteosclerotic bone has invaded the previously osteolytic regions of affected bone, evidence of ongoing abnormal bone resorption can be seen in the form of secondary resorption fronts. These secondary fronts are commonly seen as clefts in the cortex of long bones and trail in the path of the primary front.

In the final stage of Paget’s disease, termed osteoblastic or sclerotic, the affected bone remains dense and retains the “mosaic” matrix pattern characteristic of Paget’s disease. Tubular bones show a loss of differentiation between cortical and trabecular bone, even with enlargement of the bone as a result of periosteal new bone formation, because the new bone is no longer compact bone. Much less cellular activity is present in sclerotic lesions. Osteoclasts are few or absent, but osteoblasts may still be seen to line bone surfaces. The marrow may remain fibrous, but the numbers of blood vessels are greatly reduced. Scattered chronic inflammatory cells may be present. Occasionally, parts of lesions are totally devoid of bone cells, and for this reason, the concept of “burned-out” Paget’s disease has arisen. However, it is very unlikely that extensive lesions of Paget’s disease ever achieve an entirely burned-out state. On the contrary, the presence of all stages of the disease in a single bone is much more likely.

The evolution of Paget’s disease can also be observed by administration of radioactive tracers and scanning of the entire skeleton or selected regions. Bone scans use technetium-labeled bisphosphonates, which after intravenous injection localize to skeletal sites in proportion to the relative blood flow and the rate of bone formation. The scans usually, but not always, demonstrate high uptake of radioactivity in the areas of the skeleton noted to be radiographically abnormal³⁷ (Fig. 16-5). In a small proportion of patients, increased uptake may be seen when the radiograph is normal, thus illustrating the great sensitivity of bone scans. On the contrary, a small percentage of sclerotic lesions may not be picked up by a bone scan. These lesions appear to represent areas of inactive disease. Bone scans can be analyzed semiquantitatively or by computer and thus may be used to monitor response to treatment.

Gallium scans, most often used to detect occult infection or tumors, also have been shown to delineate the lesions of Paget’s disease.³⁸ Evidence indicates that tracer gallium is localized to the nuclei of osteoclasts.³⁹ Therefore, the gallium scan may serve as a direct index of cellular activity in Paget’s disease.

Clinical Features

A considerable proportion of individuals with Paget’s disease have neither symptoms nor signs of the disease.¹ The disease is accidentally discovered in these cases because of radiologic or biochemical abnormalities uncovered during investigation of another disorder.

The most common clinical problems are pain and deformity. The bone pain of Paget’s disease, when present, is seldom severe. Usually it is a dull pain, is located deep below the soft tissues, and often persists during the night. In weight-bearing bone, the pain may be slightly worsened by ambulation, but to a lesser extent than pain originating from the joints or from nerve impingement. Deformity of the skeleton is most often noted in the skull and the lower extremity. Over many years, the hat size of an affected individual may be noted to increase. Bowing and enlargement of the femur and tibia also may evolve over a period of years.