Type 1 Collagen Diseases (Osteogenesis Imperfecta)

Osteogenesis imperfecta, or brittle bone disease, is a phenotypically diverse disorder caused by deficiencies in the synthesis of type 1 collagen. It is the most common inherited disorder of connective tissue. It principally affects bone, but also impacts other tissues rich in type 1 collagen (joints, eyes, ears, skin, and teeth). Osteogenesis imperfecta usually results from autosomal dominant mutations (over 800 have been identified) in the genes that enco de theα1 and α2 chains of collagen.¹⁹ Many of these mutations involve the substitution of glycine residues in the triple-helical domain. The genotypephenotype relationship underlying osteogenesis imperfecta is based on the location of the mutation within the protein. Mutations resulting in decreased synthesis of qualitatively normal collagen are associated with mild skeletal abnormalities. More severe or lethal phenotypes have abnormal polypeptide chains that cannot be arranged in the triple helix. Recently, mutations in the genes for cartilage-associated protein (CRTAP) and leucine proline-enriched proteoglycan 1 (LEPRE1) have been shown to be responsible for three rare variants of the disease.²⁰

Clinically, osteogenesis imperfecta is separated into four major subtypes that vary widely in severity (Table 26-3). The type II variant is at one end of the spectrum and is uniformly fatal in utero or during the perinatal period. It is characterized by extraordinary bone fragility with multiple intrauterine fractures (Fig. 26-6). In contrast, individuals with the type I form have a normal life span but experience childhood fractures that decrease in frequency following puberty. Other findings include blue sclerae caused by decreased collagen content, making the sclera translucent and allowing partial visualization of the underlying choroid; hearing loss related to both a sensorineural deficit and impeded conduction due to abnormalities in the bones of the middle and inner ear; and dental imperfections (small, misshapen, and blue-yellow teeth) secondary to a deficiency in dentin. The basic abnormality in all forms of osteogenesis imperfecta is too little bone, thus constituting a type of osteoporosis with marked cortical thinning and attenuation of trabeculae.

TABLE 26-3 Subtypes of Osteogenesis Imperfecta

FIGURE 26-6 Skeletal radiogram of a fetus with lethal type II osteogenesis imperfecta. Note the numerous fractures of virtually all bones, resulting in accordion-like shortening of the limbs.

Diseases Associated with Mutations of Types 2, 9, 10, and 11 Collagen

Types 2, 9, 10, and 11 collagens are important structural components of hyaline cartilage. Although uncommon, mutations in the genes encoding them produce an array of disorders ranging from the fatal to those compatible with life but associated with early destruction of joints (see Table 26-2). In the severe disorders, the type 2 collagen molecules are not secreted by the chondrocytes, and insufficient bone formation occurs. In the milder disorders there is reduced synthesis of normal type 2 collagen.

Mucopolysaccharidoses

The mucopolysaccharidoses, as discussed earlier (Chapter 5), are a group of lysosomal storage diseases that are caused by deficiencies in the enzymes that degrade dermatan sulfate, heparan sulfate, and keratan sulfate. The affected enzymes are mainly acid hydrolases. Mesenchymal cells, especially chondrocytes, normally metabolize extracellular matrix mucopolysaccharides; hence, cartilage formation is severely affected. Consequently, many of the skeletal manifestations of the mucopolysaccharidoses result from abnormalities in hyaline cartilage, including the cartilage anlage, growth plates, costal cartilages, and articular surfaces. It is not surprising, therefore, that affected individuals are frequently of short stature and have chest wall abnormalities, and malformed bones.

Osteopetrosis

Osteopetrosis, also known as marble bone disease and Albers-Schönberg disease, refers to a group of rare genetic diseases that are characterized by reduced bone resorption and diffuse symmetric skeletal sclerosis due to impaired formation or function of osteoclasts (Fig. 26-7). The term osteopetrosis reflects the stonelike quality of the bones; however, the bones are abnormally brittle and fracture easily, like a piece of chalk. Osteopetrosis is classified into variants based on both the mode of inheritance and the clinical findings. The two major groups include autosomal recessive and dominant forms. The autosomal recessive type is further divided into mild and severe variants. The autosomal recessive severe type and the autosomal dominant mild type are the most common variants.

FIGURE 26-7 Radiogram of the upper extremity in an individual with osteopetrosis. The bones are diffusely sclerotic, and the distal metaphyses of the ulna and radius are poorly formed

(Erlenmeyer flask deformity).

Pathogenesis.

Most of the mutations underlying osteopetrosis interfere with the process of acidification of the osteoclast resorption pit, which is required for the dissolution of the calcium hydroxyapatite within the matrix. Examples include autosomal recessive defects in the gene CA2, which encodes the enzyme carbonic anhydrase II.²¹ Carbonic anhydrase II is required by osteoclasts and renal tubular cells to generate protons from carbon dioxide and water. Absence of CAII prevents osteoclasts from acidifying the resorption pit and solubilizing hydroxyapatite, and also blocks the acidification of urine by the renal tubular cells. In an autosomal recessive severe form of the disease, a mutation in the chloride channel gene CLCN7 interferes with the function of the H⁺-ATPase proton pump located on the osteoclast ruffled border.²¹ Another severe autosomal recessive form is caused by a mutation in the gene TCIRG1, which encodes a component of the proton pump. A less severe autosomal recessive variant results from a mutation in the gene that encodes RANKL. Not surprisingly, these individuals have fewer osteoclasts than normal. In animals, osteopetrosis can also be caused by mutations in a large number of other genes, including M-CSF, RANK, and OPG, which you will recall regulate osteoclast formation and function.²¹

Morphology. The morphologic changes of osteopetrosis are explained by deficient osteoclast activity. The bones lack a medullary canal, and the ends of long bones are bulbous (Erlenmeyer flask deformity) and misshapen. The neural foramina are small and compress exiting nerves. The primary spongiosa, which is normally removed during growth, persists and fills the medullary cavity, leaving no room for the hematopoietic marrow and preventing the formation of mature trabeculae (Fig. 26-8). Deposited bone is not remodeled and tends to be woven in architecture. In the end, these intrinsic abnormalities cause the bone to be brittle and predisposed to fracture. Histologically, the number of osteoclasts may be normal, increased, or decreased depending on the underlying genetic defect.

FIGURE 26-8 Section of proximal tibial diaphysis from a fetus with osteopetrosis. The cortex (1) is present, but the medullary cavity (2) is filled with primary spongiosa, which replaces the hematopoietic elements.

Clinical Features.

Severe infantile malignant osteopetrosis is autosomal recessive and usually becomes evident in utero or soon after birth. Fracture, anemia, and hydrocephaly are often seen, resulting in postpartum mortality. Affected individuals who survive into their infancy have cranial nerve defects (optic atrophy, deafness, and facial paralysis) and repeated—often fatal—infections because of inadequacies of the marrow produced in extramedullary sites, which also causes prominent hepatosplenomegaly. The mild autosomal dominant benign form may not be detected until adolescence or adulthood, when it is discovered on x-rays performed because of repeated fractures. These individuals may also have mild cranial nerve deficits and anemia.

Osteopetrosis was the first genetic disease treated with bone marrow transplantation, since osteoclasts are derived from marrow monocyte precursors. The donor progenitor cells produce normal functioning osteoclasts, which reverse many of the skeletal abnormalities.

Osteoporosis

Osteoporosis is a disease characterized by porous bones and a reduced bone mass. The associated structural changes predispose the bone to fracture. The disorder may be localized to a certain bone or region, as in disuse osteoporosis of a limb, or may involve the entire skeleton, as a manifestation of a metabolic bone disease. Generalized osteoporosis, in turn, may be primary or secondary to a large variety of conditions (Table 26-4).

TABLE 26-4 Categories of Generalized Osteoporosis

When the term osteoporosis is used in an unqualified manner, it usually refers to the most common forms, senile and postmenopausal osteoporosis, in which the loss of bone mass makes the skeleton vulnerable to fractures. It is estimated that one million Americans experience a fracture related to osteoporosis each year, at a cost of over 14 billion dollars. Effective treatment and prevention are imperative. The following discussion relates largely to these dominant forms of osteoporosis.

Pathogenesis.

Peak bone mass is achieved during young adulthood. Its magnitude is determined largely by hereditary factors, especially polymorphisms in the genes that influence bone metabolism (discussed later).²² Physical activity, muscle strength, diet, and hormonal state also make important contributions. Once maximal skeletal mass is attained, a small deficit in bone formation accrues with every resorption and formation cycle of each basic multicellular unit. Accordingly, age-related bone loss, which may average 0.7% per year, is a normal and predictable biologic phenomenon. Both sexes are affected equally and whites more so than blacks. Differences in the peak skeletal mass in men versus women and in blacks versus whites may partially explain why certain populations are prone to develop this disorder.

Although much remains unknown, discoveries in the molecular biology of bone have provided intriguing new hypotheses about the pathogenesis of osteoporosis (Fig. 26-9):

• Age-related changes in bone cells and matrix have a strong impact on bone metabolism. Osteoblasts from elderly individuals have reduced proliferative and biosynthetic potential when compared with osteoblasts from younger individuals.²³ Also, proteins bound to the extracellular matrix (such as growth factors, which are both mitogenic to osteoprogenitor cells and stimulate osteoblastic synthetic activity) lose their biologic punch over time. The net result is a diminished capacity to make bone. This form of osteoporosis, known as senile osteoporosis, is categorized as a low-turnover variant.

• Reduced physical activity increases the rate of bone loss in experimental animals and humans, because mechanical forces stimulate normal bone remodeling. Bone loss in an immobilized or paralyzed extremity, the reduction of skeletal mass in astronauts in a zero gravity environment for prolonged periods, and the higher bone density in athletes exemplify the role of physical activity in preventing bone loss. The type of exercise is important, as load magnitude influences bone density more than the number of load cycles. Because muscle contraction is the dominant source of skeletal loading, resistance exercises such as weight training are more effective stimuli for increasing bone mass than repetitive endurance activities such as jogging. Certainly the decreased physical activity that is associated with aging contributes to senile osteoporosis.

• Genetic factors are also important, as noted previously. It is estimated that 60% to 80% of the variation in bone density is genetically determined. In genome-wide association studies (GWAS), the top associated genes include RANKL, OPG, and RANK,²⁴ which you will recall encode key regulators of osteoclasts. Also associated are the MHC locus (perhaps reflecting the effects of inflammation on calcium metabolism) and the estrogen receptor gene (discussed below). Some studies have also implicated genetic variants of the vitamin D receptor and LRP5 as risk factors.

• The body’s calcium nutritional state is important. It has been shown that adolescent girls (but not boys) tend to have insufficient calcium intake in the diet. This calcium deficiency occurs during a period of rapid bone growth, stunting the peak bone mass ultimately achieved; thus, these individuals are at greater risk of developing osteoporosis. Calcium deficiency, increased PTH concentrations, and reduced levels of vitamin D may also have a role in the development of senile osteoporosis.

• Hormonal influences. In the decade after menopause, yearly reductions in bone mass may reach up to 2% of cortical bone and 9% of cancellous bone. Women may lose as much as 35% of their cortical bone and 50% of their trabecular bone within the 30 to 40 years after menopause. It is thus no surprise that post-menopausal women suffer osteoporotic fractures more commonly than men of the same age. Postmenopausal osteoporosis is characterized by a hormone-dependent acceleration of bone loss that occurs during the decade after menopause. Estrogen deficiency plays the major role in this phenomenon, and estrogen replacement at menopause is protective against bone loss. The effects of estrogen on bone mass are mediated by cytokines. Decreased estrogen levels appear to result in increased secretion of inflammatory cytokines by blood monocytes and bone marrow cells. These cytokines stimulate osteoclast recruitment and activity by increasing the levels of RANKL while diminishing the expression of OPG. Compensatory osteoblastic activity occurs, but it does not keep pace, leading to what is classified as a high-turnover form of osteoporosis.

FIGURE 26-9 Pathophysiology of postmenopausal and senile osteoporosis (see text).

Morphology. The entire skeleton is affected in postmenopausal and senile osteoporosis (Fig. 26-10), but certain regions tend to be more severely involved than others. In postmenopausal osteoporosis the increase in osteoclast activity affects mainly bones or portions of bones that have increased surface area, such as the cancellous compartment of vertebral bodies. The trabecular plates become perforated, thinned, and lose their interconnections, leading to progressive microfractures and eventual vertebral collapse. In senile osteoporosis the cortex is thinned by subperiosteal and endosteal resorption, and the haversian systems are widened. In severe cases the haversian systems are so enlarged that the cortex mimics cancellous bone. The bone that remains is of normal composition.

FIGURE 26-10 Osteoporotic vertebral body (right) shortened by compression fractures compared with a normal vertebral body. Note that the osteoporotic vertebra has a characteristic loss of horizontal trabeculae and thickened vertical trabeculae.

Clinical Course.

The clinical manifestations of osteoporosis depend on which bones are involved. Vertebral fractures that frequently occur in the thoracic and lumbar regions are painful, and when multiple can cause significant loss of height and various deformities, including lumbar lordosis and kyphoscoliosis. Complications of fractures of the femoral neck, pelvis, or spine, such as pulmonary embolism and pneumonia, are frequent and result in 40,000 to 50,000 deaths per year.

Osteoporosis cannot be reliably detected in plain radiograms until 30% to 40% of the bone mass is lost, and measurement of blood levels of calcium, phosphorus, and alkaline phosphatase are not diagnostic. Osteoporosis is thus a difficult condition to diagnose accurately, since it remains asymptomatic until skeletal fragility is well advanced. The best procedures to accurately estimate the amount of bone loss, aside from biopsy, are specialized radiographic imaging techniques, such as dual-energy X-ray absorptiometry and quantitative computed tomography, which measure bone density.

The prevention and treatment of senile and postmenopausal osteoporosis includes exercise, appropriate calcium and vitamin D intake, and pharmacologic agents, most commonly bisphosphonates, which bind to bone and inhibit osteoclasts.

Paget Disease (Osteitis Deformans)

This unique skeletal disease can be divided into three phases; (1) an initial osteolytic stage, followed by (2) a mixed osteoclastic-osteoblastic stage, which ends with a predominance of osteoblastic activity and evolves ultimately into (3) a burnt-out quiescent osteosclerotic stage (Fig. 26-11). The net effect is a gain in bone mass; however, the newly formed bone is disordered and architecturally unsound.

FIGURE 26-11 Diagrammatic representation of Paget disease of bone demonstrating the three phases in the evolution of the disease.

Paget disease usually begins in late adulthood (average age at diagnosis, 70 years) and becomes progressively more common thereafter. An intriguing aspect is the striking variation in its prevalence, both within certain countries and throughout the world. Paget disease is relatively common in whites in England, France, Austria, regions of Germany, Australia, New Zealand, and the United States. In contrast, Paget disease is rare in the native populations of Scandinavia, China, Japan, and Africa. The exact incidence is hard to determine because many affected individuals are asymptomatic; it is estimated that 1% of the US population over the age of 40 is affected and the prevalence rate in England is 2.5% for men and 1.6% for women 55 years or older. Recent surveys show that there has been a fall in new cases over the last 25–30 years, and a decline in its clinical severity.

Pathogenesis.

The cause of Paget disease remains uncertain, and current evidence suggests both environmental and genetic factors. The risk of developing the disorder is approximately seven times greater in first-degree relatives of affected individuals than it is in normal controls,²⁵ and 15% to 40% of individuals with Paget disease have a family history that shows an autosomal dominant pattern of inheritance. Mutations in the SQSTM1 gene are present in 40% to 50% of cases of familial Paget disease, and in 5% to 10% of patients without a family history. The SQSTM mutations enhance NF-κB activation by RANK signaling, leading to increased osteoclast activity and an increased susceptibility to the disease. Mutations in RANKL and RANK/OPG have also been found in genetic diseases that have some phenotypic overlap with Paget disease; including familial expansile osteolysis, expansile skeletal hyperphosphatasia, early-onset Paget disease, juvenile Paget disease, and the syndrome of hereditary inclusion body myopathy, and frontotemporal dementia.

When Sir James Paget first described this condition in 1876, he attributed the skeletal changes to an inflammatory process, hence the term osteitis deformans. Support for this idea over the years has centered on a possible role for infection by a paramyxovirus, but this hypothesis remains unproven.

Morphology. Paget disease is a focal process that shows remarkable histologic variation over time and from site to site. The hallmark is the mosaic pattern of lamellar bone. This pattern, which is likened to a jigsaw puzzle, is produced by prominent cement lines that anneal haphazardly oriented units of lamellar bone (Fig. 26-12). In the initial lytic phase there are waves of osteoclastic activity and numerous resorption pits. The osteoclasts are abnormally large and have many more than the normal 10 to 12 nuclei; sometimes 100 nuclei are present. Osteoclasts persist in the mixed phase, but now many of the bone surfaces are lined by prominent osteoblasts. The marrow adjacent to the bone-forming surface is replaced by loose connective tissue that contains osteoprogenitor cells and numerous blood vessels, which transport nutrients and catabolites to and from these metabolically active sites. The newly formed bone may be woven or lamellar, but eventually all of it is remodeled into lamellar bone. As the mosaic pattern unfolds and the cell activity decreases, the periosseous fibrovascular tissue recedes and is replaced by normal marrow. In the end, the bone becomes a caricature of itself: larger than normal and composed of coarsely thickened trabeculae and cortices that are soft and porous and lack structural stability. These aspects make the bone vulnerable to deformation under stress; consequently, it fractures easily.

FIGURE 26-12 Mosaic pattern of lamellar bone pathognomonic of Paget disease.

Clinical Course.

Clinical findings are extremely variable and depend on the extent and site of the disease. Most cases are mild and are discovered as an incidental radiographic finding. Paget disease is monostotic in about 15% of cases and polyostotic in the remainder. The axial skeleton or proximal femur is involved in up to 80% of cases. Even though no bone is immune, involvement of the ribs, fibula, and small bones of the hands and feet is unusual.

Pain localized to the affected bone is common. It is caused by microfractures or by bone overgrowth that compresses spinal and cranial nerve roots. Enlargement of the craniofacial skeleton may produce leontiasis ossea and a cranium so heavy that is difficult for the person to hold the head erect. The weakened pagetic bone may lead to invagination of the skull base (platybasia) and compression of the posterior fossa structures. Weight bearing causes anterior bowing of the femurs and tibiae and distorts the femoral heads, resulting in the development of severe secondary osteoarthritis. Chalkstick-type fractures are another frequent complication and usually occur in the long bones of the lower extremities. Compression fractures of the spine result in spinal cord injury and the development of kyphoses. The hypervascularity of pagetic bone warms the overlying skin, and in severe polyostotic disease the increased blood flow acts like an arteriovenous shunt, leading to high-output heart failure or exacerbation of underlying cardiac disease.

A variety of tumor and tumor-like conditions develop in pagetic bone. The benign lesions include giant-cell tumor, giant-cell reparative granuloma, and extra-osseous masses of hematopoiesis. The most dreaded complication is sarcoma, which occurs in 0.7% to 0.9% of all individuals with Paget disease, and in 5% to 10% of those with severe polyostotic disease. The sarcomas are usually osteosarcoma or fibrosarcoma, and they arise in Paget lesions in the long bones, pelvis, skull, and spine.

The diagnosis can frequently be made from the radiographic findings. Pagetic bone is typically enlarged with thick, coarsened cortices and cancellous bone(Fig. 26-13). Active disease has a wedge-shaped lytic leading edge that may progress along the length of the bone at a rate of 1 cm per year.²⁶ Many affected individuals have elevated serum alkaline phosphatase levels and increased urinary excretion of hydroxyproline.

FIGURE 26-13 Severe Paget disease. The tibia is bowed and the affected portion is enlarged, sclerotic, and exhibits irregular thickening of both the cortical and cancellous bone.

In the absence of malignant transformation, Paget disease is usually not a serious or life-threatening disease. Most affected individuals have mild symptoms that are readily suppressed by calcitonin and bisphosphonates.

Rickets and Osteomalacia

Rickets and osteomalacia are disorders characterized by a defect in matrix mineralization, most often related to a lack of vitamin D or some disturbance in its metabolism. The term rickets refers to the disorder in children in which deranged bone growth produces distinctive skeletal deformities. In the adult the disorder is called osteomalacia, because the bone that forms during the remodeling process is inadequately mineralized. This results in osteopenia and predisposition to insufficiency fractures. Both rickets and osteomalacia are discussed in Chapter 9.

Hyperparathyroidism

Hyperparathyroidism is classified into primary and secondary types as discussed in Chapter 24. Primary hyperparathyroidism results from autonomous hyperplasia or a tumor, usually an adenoma, of the parathyroid gland, whereas secondary hyperparathyroidism is commonly caused by prolonged states of hypocalcemia resulting in compensatory hypersecretion of PTH. Whatever the basis, the increased PTH concentrations are detected by receptors on osteoblasts, which then release factors that stimulate osteoclast activity. Thus, through a chain of signals, the skeletal manifestations of hyperparathyroidism are caused by unabated osteoclastic bone resorption. The following points should be noted:

Morphology. For unknown reasons, the increased osteoclast activity in hyperparathyroidism affects cortical bone (subperiosteal, osteonal, and endosteal surfaces) more severely than cancellous bone. Subperiosteal resorption produces thinned cortices and the loss of the lamina dura around the teeth. X-rays reveal a pattern of radiolucency that is virtually diagnostic of hyperparathyroidism. In cancellous bone, osteoclasts tunnel into and dissect centrally along the length of the trabeculae, creating the appearance of railroad tracks and producing what is known as dissecting osteitis (Fig. 26-14). The correlative radiographic finding is a decrease in bone density or osteopenia. Since bone resorption and formation are coupled processes, it is not surprising that osteoblast activity is also increased in hyperparathyroidism. The marrow spaces around the affected surfaces are replaced by fibrovascular tissue.

FIGURE 26-14 Hyperparathyroidism with osteoclasts boring into the center of the trabeculum (dissecting osteitis).

The bone loss predisposes to microfractures and secondary hemorrhages that elicit an influx of macrophages and an ingrowth of reparative fibrous tissue, creating a mass of reactive tissue, known as a brown tumor (Fig. 26-15). The brown color is the result of the vascularity, hemorrhage, and hemosiderin deposition, and it is not uncommon for the lesions to undergo cystic degeneration. The combined picture of increased bone cell activity, peritrabecular fibrosis, and cystic brown tumors is the hallmark of severe hyperparathyroidism and is known as generalized osteitis fibrosa cystica (von Recklinghausen disease of bone).

FIGURE 26-15 Resected rib, harboring an expansile brown tumor adjacent to the costal cartilage.

The decrease in bone mass predisposes to fractures, deformities caused by the stress of weight bearing, and joint pain and dysfunction as the lines of normal weight bearing are altered. Control of the hyperparathyroidism allows the bony changes to regress significantly or disappear completely.

Renal Osteodystrophy

The term renal osteodystrophy is used to describe collectively all of the skeletal changes of chronic renal disease, including (1) increased osteoclastic bone resorption mimicking osteitis fibrosa cystica, (2) delayed matrix mineralization (osteomalacia), (3) osteosclerosis, (4) growth retardation, and (5) osteoporosis. As advances in medical technology have prolonged the lives of individuals with renal disease, its impact on skeletal homeostasis has assumed greater clinical importance.

The various histologic bone changes in individuals with end-stage renal failure can be divided into three major types of disorders.²⁷ High-turnover osteodystrophy is characterized by increased bone resorption and bone formation, with the former predominating. In contrast, low-turnover or aplastic disease is manifested by adynamic bone (little osteoclastic and osteoblastic activity) and, less commonly, osteomalacia. Many affected individuals have the third type, which is a mixed pattern of disease.

Osteoma

Osteomas are bosselated, round-to-oval sessile tumors that project from the subperiosteal surface of the cortex. They most often arise on or inside the skull and facial bones. They are usually solitary and are detected in middle age. Multiple osteomas are seen in the setting of Gardner syndrome (Chapter 17). They consist of a composite of woven and lamellar bone that is frequently deposited in a cortical pattern with haversian-like systems. Some variants contain a component of trabecular bone in which the intertrabecular spaces are filled with hematopoietic marrow.

Osteomas are generally slow-growing tumors of little clinical significance except when they cause obstruction of a sinus cavity, impinge on the brain or eye, interfere with function of the oral cavity, or produce cosmetic problems.

Osteoid Osteoma and Osteoblastoma

Osteoid osteoma and osteoblastoma are terms used to describe benign bone tumors that have identical histologic features but differ in size, sites of origin, and symptoms. Osteoid osteomas are by definition less than 2 cm in greatest dimension and usually occur in the teens and 20s. Seventy-five percent of affected individuals are younger than 25 years old, and men outnumber women 2 : 1. They can arise in any bone but have a predilection for the appendicular skeleton and posterior elements of the spine. In 50% of cases the femur or tibia is involved, wherein they commonly arise in the cortex and less frequently within the medullary cavity. Osteoid osteomas produce severe nocturnal pain that is relieved by aspirin.³³ The pain is probably caused by excess prostaglandin E₂ (PGE₂) production by the proliferating osteoblasts. Osteoblastoma is larger than 2 cm and involves the spine more frequently; the pain is dull, achy, and unresponsive to salicylates, and the tumor usually does not induce a marked bony reaction.

Morphology. Osteoid osteoma and osteoblastoma are round-to-oval masses of hemorrhagic gritty tan tissue. They are well circumscribed and composed of randomly interconnecting trabeculae of woven bone that are prominently rimmed by osteoblasts (Fig. 26-19). The stroma surrounding the neoplastic bone consists of loose connective tissue that contains many dilated and congested capillaries. The relatively small size, well-defined margins, and benign cytologic features of the neoplastic osteoblasts help distinguish these tumors from osteosarcoma. Osteoid osteomas, especially those that arise beneath the periosteum, usually elicit a tremendous amount of reactive bone formation that encircles the lesion. The actual tumor, known as the nidus, manifests radiographically as a small round lucency that may be centrally mineralized (Fig. 26-20).

FIGURE 26-19 Osteoid osteoma composed of haphazardly interconnecting trabeculae of woven bone that are rimmed by prominent osteoblasts. The intertrabecular spaces are filled by vascularized loose connective tissue.

FIGURE 26-20 Specimen radiogram of intracortical osteoid osteoma. The round radiolucency with central mineralization represents the lesion and is surrounded by abundant reactive bone that has massively thickened the cortex.

Osteoid osteoma is frequently treated by radioablation. Osteoblastoma is usually curetted or excised en bloc in a conservative fashion. The possibility of malignant transformation is remote except when osteoblastoma is treated with radiation (large tumors in the base of the skull and spine), which may promote this dreaded complication.

Osteosarcoma

Osteosarcoma is a malignant mesenchymal tumor in which the cancerous cells produce bone matrix. It is the most common primary malignant tumor of bone, exclusive of myeloma and lymphoma, and accounts for approximately 20% of primary bone cancers. Osteosarcoma occurs in all age groups but has a bimodal age distribution; 75% occur in persons younger than 20 years of age.³⁴ The smaller second peak occurs in the elderly, who frequently suffer from conditions known to predispose to osteosarcoma—Paget disease, bone infarcts, and prior irradiation. Overall, men are more commonly affected than women (1.6 : 1). The tumors usually arise in the metaphyseal region of the long bones of the extremities, and almost 50% occur about the knee (Fig. 26-21). Any bone can be involved, however, and in persons beyond the age of 25, the incidence in flat bones and long bones is almost equal.

FIGURE 26-21 Major sites of origin of osteosarcomas. The numbers are approximate percentages arising at each site.

Pathogenesis.

Approximately 70% of osteosarcomas have acquired genetic abnormalities such as ploidy changes and chromosomal aberrations, none of which are specific to this tumor. More telling is the presence of very frequent mutations that interfere with function of two genes: (1) RB, the retinoblastoma gene, a critical cell cycle regulator; and (2) p53, a gene whose product regulates DNA repair and certain aspects of cellular metabolism (Chapter 7). Although the basic mechanisms that cause the development of osteosarcoma are still unknown, it is clear that defects in RB and p53 play important roles in the process. This association is emphasized by rare patients with germline mutations in RB, who have a roughly 1000-fold increased risk of osteosarcoma; and similarly by patients with Li-Fraumeni syndrome (germline p53 mutations), who also have a greatly elevated incidence of this tumor. Abnormalities in INK4a, which encodes p16 (a cell cycle regulator) and p14 (which aids and abets p53 function), also are seen in osteosarcoma. It is also noteworthy that osteosarcomas tend to occur at sites of bone growth, presumably because proliferation makes osteoblastic cells prone to acquire mutations that could lead to transformation. The association may contribute to the high incidence of osteosarcoma in large dog breeds, such as St. Bernards and Great Danes.

Morphology. Several subtypes of osteosarcoma are recognized and are grouped according to

• Site of origin (intramedullary, intracortical, or surface)

• Degree of differentiation

• Multicentricity (synchronous, metachronous)

• Primary (underlying bone is unremarkable) or secondary to preexisting disorders such as benign tumors, Paget disease, bone infarcts, previous irradiation

• Histologic features (osteoblastic, chondroblastic, fibroblastic, telangiectatic, small cell, and giant cell).

The most common subtype arises in the metaphysis of long bones and is primary, solitary, intramedullary, and poorly differentiated.

Grossly, osteosarcomas are big bulky tumors that are gritty, gray-white, and often contain areas of hemorrhage and cystic degeneration (Fig. 26-22). The tumors frequently destroy the surrounding cortices and produce soft-tissue masses. They spread extensively in the medullary canal, infiltrating and replacing the marrow surrounding the preexisting bone trabeculae. Infrequently, they penetrate the epiphyseal plate or enter the joint. When joint invasion occurs, the tumor grows into it along tendoligamentous structures or through the attachment site of the joint capsule. The tumor cells vary in size and shape and frequently have large hyperchromatic nuclei. Bizarre tumor giant cells are common, as are mitoses. The formation of bone by the tumor cells is characteristic (Fig. 26-23). The neoplastic bone usually hasa coarse, lace-like architecture but also may be deposited in broad sheets or as primitive trabeculae. Other matrices, including cartilage or fibrous tissue, may be present in varying amounts. When malignant cartilage is abundant, the tumor is called chondroblastic osteosarcoma. Vascular invasion is usually conspicuous, and up to 50% to 60% of an individual tumor may be necrotic.

FIGURE 26-22 Osteosarcoma of the upper end of the tibia. The tan-white tumor fills most of the medullary cavity of the metaphysis and proximal diaphysis. It has infiltrated through the cortex, lifted the periosteum, and formed soft-tissue masses on both sides of the bone.

FIGURE 26-23 Coarse, lacelike pattern of neoplastic bone produced by anaplastic malignant tumor cells. Note the mitotic figures.

Clinical Course.

Osteosarcomas typically present as painful, progressively enlarging masses. Sometimes a sudden fracture of the bone is the first symptom. Radiograms of the primary tumor usually show a large destructive, mixed lytic and blastic mass with infiltrative margins (Fig. 26-24). The tumor frequently breaks through the cortex and lifts the periosteum, resulting in reactive periosteal bone formation. The triangular shadow between the cortex and raised ends of periosteum is known radiographically as Codman triangle and is characteristic but not diagnostic of this tumor. These aggressive neoplasms spread hematogenously, and at the time of diagnosis approximately 10% to 20% of affected individuals have demonstrable pulmonary metastases, and it is likely that many more have occult metastases. In those who die of the neoplasm, 90% have metastases to the lungs, bones, brain, and elsewhere.

FIGURE 26-24 Distal femoral osteosarcoma with prominent bone formation extending into the soft tissues. The periosteum, which has been lifted, has laid down a proximal triangular shell of reactive bone known as a Codman triangle (arrow).

Osteosarcoma is treated with a multimodality approach that includes chemotherapy, which is given under the assumption that all patients at the time of diagnosis have metastases, which are usually too small to detect by imaging. The prognosis of patients without detectable metastases has improved substantially, with 5-year survival rates reaching 60% to 70% with aggressive chemotherapy and limb salvaging surgery. Unfortunately, the outcome for patients with overt metastases or recurrent disease is still poor (approximately 20% 5-year survival rate).

Osteochondroma

Osteochondroma, also known as an exostosis, is a benign cartilage-capped tumor that is attached to the underlying skeleton by a bony stalk. It is the most common benign bone tumor; about 85% are solitary. The remainder are seen as part of the multiple hereditary exostosis syndrome, which is an autosomal dominant hereditary disease. Hereditary exostoses are caused by germline loss-of-function mutations in either the EXT1 or EXT2 genes, whereas inactivation of only EXT1 has been detected in sporadic tumors. These genes encode proteins that function in the biosynthesis of heparin sulfate proteoglycans (Chapter 3). Reduced expression of EXT1 and EXT2 results in defective endochondral ossification, which somehow sets the stage for abnormal growth. Solitary osteochondromas are usually first diagnosed in late adolescence and early adulthood, but multiple osteochondromas become apparent during childhood. Men are affected three times more often than women. Osteochondromas develop only in bones of endochondral origin and arise from the metaphysis near the growth plate of long tubular bones, especially about the knee. Occasionally, they develop from bones of the pelvis, scapula, and ribs, and in these sites they are frequently sessile and have short stalks. Rarely, they involve the short tubular bones of the hands and feet.

Morphology. Osteochondromas are sessile or mushroom shaped, and range in size from 1 to 20 cm. The cap is composed of benign hyaline cartilage varying in thickness (Fig. 26-25) and is covered peripherally by perichondrium. The cartilage has the appearance of disorganized growth plate and undergoes enchondral ossification, with the newly made bone forming the inner portion of the head and stalk. The cortex of the stalk merges with the cortex of the host bone, so that the medullary cavity of the osteochondroma and bone are in continuity.

FIGURE 26-25 Osteochondroma. A, X-ray of an osteochondroma arising off the posterior surface of the tibia. B, Axial CT scan shows continuity of the cortex of the bone and the center of the osteochondroma. The fibula is adjacent to the mass. C, Gross specimen of sessile osteochondroma composed of a cap of hyaline cartilage undergoing enchondral ossification. D, The cartilage cap has the histologic appearance of disorganized growth plate-like cartilage.

Clinically, osteochondromas present as slow-growing masses, which can be painful if they impinge on a nerve or if the stalk is fractured. In many cases they are detected as an incidental finding. In multiple hereditary exostosis the underlying bones may be bowed and shortened, reflecting an associated disturbance in epiphyseal growth. Osteochondromas usually stop growing at the time of growth plate closure. Rarely in sporadic cases, but more commonly in those with multiple hereditary exostosis, they give rise to a chondrosarcoma or some other type of sarcoma.

Chondromas

Chondromas are benign tumors of hyaline cartilage that usually occur in bones of enchondral origin. They can arise within the medullary cavity, where they are known as enchondromas, or on the surface of bone, where they are called subperiosteal or juxtacortical chondromas. Enchondromas are the most common of the intraosseous cartilage tumors and are usually diagnosed in individuals who are in their 20s to 40s. They are usually solitary metaphyseal lesions of tubular bones; the favored sites are the short tubular bones of the hands and feet. A syndrome of multiple enchondromas or enchondromatosis is known as Ollier disease. If the enchondromatosis is associated with soft-tissue hemangiomas, the disorder is called Maffucci syndrome.

Morphology. Enchondromas are usually smaller than 3 cm and grossly are gray-blue and translucent. They are composed of well-circumscribed nodules of cyto logically benign hyaline cartilage (Fig. 26-26). The peripheral portion of the nodules may undergo enchondral ossification, and the center can calcify and die. The chondromas in Ollier disease and Maffucci syndrome are sometimes more cellular and exhibit cytologic atypia, making it difficult to distinguish them from chondrosarcoma.

FIGURE 26-26 Enchondroma with a nodule of hyaline cartilage encased by a thin layer of reactive bone.

Clinical Features.

Most enchondromas are asymptomatic and are detected incidentally. Occasionally they are painful and cause pathologic fracture. The tumors in enchondromatosis may be numerous and large, producing severe deformities. The radiographic features are characteristic; the unmineralized nodules of cartilage produce well-circumscribed oval lucencies that are surrounded by a thin rim of radiodense bone (C or O ring sign). If the matrix calcifies it is detected as irregular opacities. The nodules scallop the endosteum, but usually leave the cortex intact (Fig. 26-27). The growth potential of chondromas is limited, and most remain stable. Treatment depends on the clinical situation and is usually observation or curettage. Solitary chondromas rarely undergo sarcomatous transformation, but those associated with enchondromatoses do so more frequently. Individuals with Maffucci syndrome are also at risk of developing other types of malignancies, including ovarian carcinomas and brain gliomas.

FIGURE 26-27 Enchondroma of the phalanx with a pathologic fracture. The radiolucent nodules of hyaline cartilage scallop the endosteal surface.

Chondroblastoma

Chondroblastoma is a rare benign tumor that accounts for less than 1% of primary bone tumors. It usually occurs in young patients in their teens and has a male-to-female ratio of 2 : 1. Most arise about the knee; less common sites such as the pelvis and ribs are affected in older patients. Chondroblastoma has a striking predilection for epiphyses and apophyses (epiphyseal equivalents, i.e., iliac crest).³⁶

Morphology. The tumor is composed of sheets of compact polyhedral chondroblasts that have well-defined cytoplasmic borders, moderate amounts of pink cytoplasm, and nuclei that are hyperlobulatedwith longitudinal grooves (Fig. 26-28). Mitotic activity and necrosis are frequently present. The tumor cells are surrounded by scant amounts of hyaline matrix that is deposited in a lace-like configuration; nodules of well-formed hyaline cartilage are distinctly uncommon. When the matrix calcifies it produces a characteristic chicken-wire pattern of mineralization (see Fig. 26-28). Scattered through the lesion are non-neoplastic osteoclast-type giant cells. Occasionally the tumors undergo prominent hemorrhagic cystic degeneration.

FIGURE 26-28 Chondroblastoma with scant mineralized matrix surrounding chondroblasts in a chicken wire–like fashion.

Chondroblastomas are usually painful, and because of their location near a joint they also cause effusions and restrict joint mobility. Radiographically, they produce a well-defined geographic lucency that commonly has spotty calcifications. Recurrences are not uncommon after curettage. Pulmonary metastases occur rarely in lesions that have undergone prior pathologic fracture or repeated curettage. Apparently in these circumstances the tumor cells are pushed into ruptured vessels, giving them access to the systemic circulation.

Chondromyxoid Fibroma

Chondromyxoid fibroma is the rarest of cartilage tumors and because of its varied morphology can be mistaken for sarcoma. It affects individuals in their teens and 20s and has a male preponderance. The tumors most frequently arise in the metaphysis of long tubular bones, but can involve virtually any bone of the body.

Morphology. The tumors range from 3 to 8 cm in greatest dimension and are well-circumscribed, solid, and glistening tan-gray. Microscopically, there are nodules of poorly formed hyaline cartilage and myxoid tissue delineated by fibrous septae. The cellularity varies; the areas of greatest cellularity are at the periphery of the nodules. In the cartilaginous regions the tumor cells are situated in lacunae; however, in the myxoid areas, the cells are stellate, and their delicate cell processes extend through the mucinous ground substance and approach or contact neighboring cells (Fig. 26-29). In contrast to other benign cartilage tumors, the neoplastic cells in chondromyxoid fibroma show varying degrees of cytologic atypia, including the presence of large hyperchromatic nuclei. Other findings include small foci of calcification of the cartilaginous matrix and scattered non-neoplastic, osteoclast-type giant cells.

FIGURE 26-29 Chondromyxoid fibroma with prominent stellate and spindle cells surrounded by myxoid matrix. Occasional osteoclast-type giant cells are also present.

Individuals with chondromyxoid fibroma usually complain of localized dull, achy pain. In most instances, radiograms demonstrate an eccentric geographic lucency that is well delineated from the adjacent bone by a rim of sclerosis. Occasionally the tumor expands the overlying cortex. The treatment of choice is simple curettage, and even though they may recur, they do not pose a threat for malignant transformation or metastasis.

Chondrosarcoma

Chondrosarcomas are a group of tumors that span a broad spectrum of clinical and pathologic findings. The feature common to all of them is the production of neoplastic cartilage. Chondrosarcoma is subclassified according to site as central (intramedullary) and peripheral (juxtacortical and surface). Histologically, they include conventional (hyaline and/or myxoid), clear cell, dedifferentiated, and mesenchymal variants. Conventional central tumors constitute about 90% of chondrosarcomas.

Chondrosarcoma of the skeleton is about half as frequent as osteosarcoma and is the second most common malignant matrix-producing tumor of bone. Individuals with chondrosarcoma are usually in their 40s or older. The clear cell and especially the mesenchymal variants occur in younger patients, in their teens or 20s. The tumor affects men twice as frequently as women. About 15% of conventional chondrosarcomas (usually peripheral tumors) arise from a preexisting enchondroma or osteochondroma.

Morphology. Conventional chondrosarcoma is composed of malignant hyaline and myxoid cartilage. The large bulky tumors are made up of nodules of gray-white, somewhat translucent glistening tissue (Fig. 26-30). In predominantly myxoid variants, the tumors are viscous and gelatinous and the matrix oozes from the cut surface. Spotty calcifications are typically present, and central necrosis may create cystic spaces. The adjacent cortex is thickened or eroded, and the tumor grows with broad pushing fronts into the surrounding soft tissue. The malignant cartilage infiltrates the marrow space and surrounds pre-existing bony trabeculae. The tumors vary in degree of cellularity, cytologic atypia, and mitotic activity (Fig. 26-31). Low-grade or grade 1 lesions demonstrate mild hypercellularity, and the chondrocytes have plump vesicular nuclei with small nucleoli. Binucleate cells are sparse, and mitotic figures are difficult to find. Portions of the matrix frequently mineralize, and the cartilage may undergo endochondral ossification. By contrast, grade 3 chondrosarcomas are characterized by marked hypercellularity, extreme pleomorphism with bizarre tumor giant cells, and mitoses. Pure grade 3 chondrosarcomas are uncommon. Such malignant cartilage is more frequently a component of chondroblastic osteosarcoma (see earlier).

FIGURE 26-30 Chondrosarcoma with lobules of hyaline and myxoid cartilage permeating throughout the medullary cavity, growing through the cortex, and forming a relatively well-circumscribed soft-tissue mass.

FIGURE 26-31 Anaplastic chondrocytes within a chondrosarcoma.

Approximately 10% of conventional low-grade chondrosarcomas have a second high-grade component that has the morphology of a poorly differentiated sarcoma; this combination defines dedifferentiated chondrosarcomas. The hallmark of clear cell chondrosarcoma is sheets of large malignant chondrocytes that have abundant clear cytoplasm, numerous osteoclast-type giant cells, and intralesional reactive bone formation, which often causes confusion with osteosarcoma. Mesenchymal chondrosarcoma is composed of islands of well-differentiated hyaline cartilage surrounded by sheets of small round cells, which can mimic Ewing sarcoma.

Chondrosarcomas commonly arise in the central portions of the skeleton, including the pelvis, shoulder, and ribs. The clear cell variant is unique in that it originates in the epiphyses of long tubular bones. In contrast to enchondroma, chondrosarcoma rarely involves the distal extremities. These tumors usually present as painful, progressively enlarging masses. The nodular growth pattern of the cartilage produces prominent endosteal scalloping radiographically. The calcified matrix appears as foci of flocculent densities. A slow-growing, low-grade tumor causes reactive thickening of the cortex, whereas a more aggressive high-grade neoplasm destroys the cortex and forms a soft-tissue mass. There is a direct correlation between the grade and the biologic behavior of the tumor.³⁷ Fortunately, most conventional chondrosarcomas are indolent and fall into the range of grade 1 and grade 2. In one analysis, the 5-year survival rates were 90%, 81%, and 43% for grades 1 through 3, respectively. None of the grade 1 tumors metastasized, whereas 70% of the grade 3 tumors disseminated. Another prognostic feature is size; tumors greater than 10 cm are more aggressive than smaller tumors. When chondrosarcomas metastasize, they spread preferentially to the lungs and skeleton. The treatment of conventional chondrosarcoma is wide surgical excision. The mesenchymal and dedifferentiated tumors are also treated with chemotherapy, because of their aggressive clinical course.

Fibrous Cortical Defect and Non-Ossifying Fibroma

Fibrous cortical defects are extremely common, being found in 30% to 50% of children older than 2 years. They are believed to be developmental defects rather than neoplasms. The vast majority arise eccentrically in the metaphysis of the distal femur and proximal tibia, and almost half are bilateral or multiple. Often they are small, about 0.5 cm in diameter. Those that grow to 5 or 6 cm in size develop into non-ossifying fibromas, which are usually not detected until adolescence.

Morphology. Both fibrous cortical defects and non-ossifying fibromas produce elongated, sharply demarcated radiolucencies that are surrounded by a thin rim of sclerosis (Fig. 26-32). They consist of gray to yellow-brown cellular lesions containing fibroblasts and macrophages (histiocytes). The cytologically bland fibroblasts are frequently arranged in a storiform (pinwheel) pattern, and the histiocytes are either multinucleated giant cells or clusters of foamy macrophages (Fig. 26-33).

FIGURE 26-32 Non-ossifying fibroma of the distal tibial metaphysis producing an eccentric lobulated radiolucency surrounded by a sclerotic margin.

FIGURE 26-33 Storiform pattern created by benign spindle cells with scattered osteoclast-type giant cells characteristic of a fibrous cortical defect and non-ossifying fibroma.

Fibrous cortical defects are asymptomatic and are usually detected on radiography as an incidental finding. Most have limited growth potential and undergo spontaneous resolution within several years, being replaced by normal cortical bone. The few that progressively enlarge into non-ossifying fibromas may present with pathologic fracture or require biopsy and curettage to exclude other types of tumors.

Fibrous Dysplasia

Fibrous dysplasia is a benign tumor that has been likened to a localized developmental arrest; all of the components of normal bone are present, but they do not differentiate into their mature structures. The lesions arise during skeletal growth and development, and appear in three distinctive but sometimes overlapping clinical patterns: (1) involvement of a single bone (monostotic); (2) involvement of multiple bones (polyostotic); and (3) polyostotic disease, associated with café-au-lait skin pigmentations and endocrine abnormalities, especially precocious puberty. The skeletal, skin, and endocrine lesions result from a somatic gain-of-function mutation occurring during embryogenesis in the GNAS gene, which you will recall is also mutated in pituitary adenomas (Chapter 24). The result of the mutations in both types of tumors is the same—the production of a hyperactive guanyl nucleotide binding protein, encoded by the GNAS gene, that drives abnormal growth.³⁸

Monostotic fibrous dysplasia accounts for 70% of all cases. It occurs equally in boys and girls, usually in early adolescence, and often stops enlarging at the time of growth plate closure. The femur, tibia, ribs, jawbones, calvaria, and humerus are most commonly affected. The lesion is frequently asymptomatic and usually discovered incidentally but it may cause pain, fracture, and discrepancies in limb length. Fibrous dysplasia can cause marked enlargement and distortion of bone, so that if the craniofacial skeleton is involved, disfigurement, sometimes severe, can occur. Monostotic disease does not evolve into the polyostotic form.

Polyostotic fibrous dysplasia without endocrine dysfunction accounts for 27% of all cases. It manifests at a slightly earlier age than the monostotic type and may continue to cause problems into adulthood. The bones affected, in descending order of frequency, are the femur, skull, tibia, humerus, ribs, fibula, radius, ulna, mandible, and vertebrae. Craniofacial involvement is present in 50% of those who have a moderate number of bones affected and in 100% of those with extensive skeletal disease. Polyostotic disease has a propensity to involve the shoulder and pelvic girdles, resulting in severe, sometimes crippling deformities (e.g., shepherd-crook deformity of the proximal femur) and spontaneous and often recurrent fractures.

Polyostotic fibrous dysplasia associated with café-au-lait skin pigmentation and endocrinopathies is known as the McCune-Albright syndrome and accounts for 3% of all cases. The endocrinopathies include sexual precocity, hyperthyroidism, pituitary adenomas that secrete growth hormone, and primary adrenal hyperplasia. The severity of manifestations in McCune-Albright syndrome depends on the number and cell types that harbor the mutation in the GNAS gene. The most common clinical presentation is precocious sexual development, which occurs most often in girls. The bone lesions are often unilateral but can be bilateral, and the skin pigmentation is usually limited to the same side of the body. The cutaneous macules are classically large; are dark to café-au-lait; have irregular serpiginous borders (coastline of Maine); and are found primarily on the neck, chest, back, shoulder, and pelvic region.

Morphology. The lesions of fibrous dysplasia are well circumscribed, intramedullary, and vary greatly in size. Larger lesions expand and distort the bone. The lesional tissue is tan-white and gritty and is composed of curvilinear trabeculae of woven bone surrounded by a moderately cellular fibroblastic proliferation. The shapes of the trabeculae mimic Chinese letters, and the bone lacks prominent osteoblastic rimming (Fig. 26-34). Nodules of hyaline cartilage with the appearance of disorganized growth plate are also present in approximately 20% of cases. Cystic degeneration, hemorrhage, and foamy macrophages are other common findings.

FIGURE 26-34 Fibrous dysplasia composed of curvilinear trabeculae of woven bone that lack conspicuous osteoblastic rimming and arise in a background of fibrous tissue.

Fibrosarcoma Variants

Collagen-producing sarcomas with a fibroblastic phenotype occur at any age, but most affect the middle-aged and elderly. They have a nearly equal sex distribution and usually arise de novo; however, a few develop in preexisting benign tumors, bone infarcts, pagetic bone, and previously irradiated tissue.

Fibrosarcoma presents as an enlarging painful mass that usually arises in the metaphysis of long bones and pelvic flat bones. Pathologic fracture is a frequent complication. Radiographically it is permeative and lytic and often extends into the adjacent soft tissue. The prognosis depends on the size, location, stage, and grade of the tumor; large, high-grade tumors that are difficult to resect have a very poor prognosis.

Ankylosing Spondyloarthritis

Also known as rheumatoid spondylitis and Marie-Strümpell disease, ankylosing spondyloarthritis is a chronic synovitis that causes destruction of articular cartilage and resultant bony ankylosis, especially of the sacroiliac and apophyseal joints (between tuberosities and processes). Inflammation of tendinoligamentous insertion sites eventuates in their ossification, producing squaring and fusion of the vertebral bodies, and bony outgrowths, which together result in severe spinal immobility. It usually becomes symptomatic in the second and third decades of life, and men are affected two to three times more frequently than women. Affected individuals characteristically present with low back pain, which frequently follows a chronic progressive course. Involvement of peripheral joints, such as the hips, knees, and shoulders, occurs in at least one third of affected individuals. Fracture of the spine, uveitis, aortitis, and amyloidosis are other recognized complications. Approximately 90% of the risk of developing the disease and the severity of the clinical manifestations is determined genetically. Although 90% of affected individuals are HLA-B27 positive, it has been suspected the other genes also contribute. Recent genome-wide association studies have shown associations with ARTS1, a gene that encodes a peptidase that trims antigens being processed for presentation by class I HLA molecules; and IL23R, the gene for the IL-23 receptor, suggesting that IL-23 (which promotes T_H17 responses) may have a role in this disease.⁴⁹

Reiter Syndrome

Reiter syndrome is a form of reactive arthritis and is defined by a triad of arthritis, nongonococcal urethritis or cervicitis, and conjunctivitis. Most affected individuals are men in their 20s or 30s, and more than 80% are HLA-B27 positive. This form of arthritis also affects individuals infected with the human immunodeficiency virus (HIV). The disease is probably caused by an autoimmune reaction initiated by prior infection of the gastrointestinal tract (Shigella, Salmonella, Yersinia, Campylobacter) and the genitourinary system (Chlamydia). Arthritic symptoms usually develop within several weeks of the inciting bout of urethritis or diarrhea. Joint stiffness and low back pain are common early symptoms. The ankles, knees, and feet are affected most often, frequently in an asymmetric pattern. Synovitis of a digital tendon sheath produces the sausage finger or toe, and ossification of tendoligamentous insertion sites leads to calcaneal spurs and bony outgrowths. Patients with severe chronic disease have involvement of the spine that is indistinguishable from ankylosing spondylitis. Extra-articular involvement manifests as inflammatory balanitis, conjunctivitis, cardiac conduction abnormalities, and aortic regurgitation. The natural behavior of Reiter syndrome is extremely variable. The episodes of arthritis usually wax and wane over a period of several weeks to 6 months. Almost 50% of affected individuals have recurrent arthritis, tendinitis, fasciitis, and lumbosacral pain that can cause significant functional disability.

Enteritis-Associated Arthritis

Enteritis-associated arthritis is caused by gastrointestinal infection by Yersinia, Salmonella, Shigella, and Campylobacter, among others. The outer cell membranes of these organisms have lipopolysaccharides as a major component, and they stimulate a host of immunological responses. The arthritis appears abruptly and tends to involve the knees and ankles but sometimes also the wrists, fingers, and toes. It lasts for about a year, then generally clears and only rarely is accompanied by ankylosing spondylitis.

Psoriatic Arthritis

Psoriatic arthritis is a chronic inflammatory arthropathy that affects peripheral and axial joints and entheses and is associated with psoriasis. Susceptibility to the disease is genetically determined and related to HLA-B27 and HLA-Cw6 alleles. It develops in more than 10% of the psoriatic population and has assorted phenotypic subtypes. Symptoms manifest between the ages of 30 and 50, and those involving the joints usually develop slowly but are acute in onset in one third of affected individuals. The patterns of joint involvement are diverse. The distal interphalangeal joints of the hands and feet are first affected in an asymmetric distribution in more than 50% of patients and may be associated with a sausage-like finger. The large joints such as the ankles, knees, hips, and wrists may be involved as well.⁵⁰ Sacroiliac and spinal disease occurs in 20% to 40% of affected individuals. Aside from conjunctivitis and iritis, extra-articular manifestations are uncommon. Histologically, psoriatic arthritis is similar to rheumatoid arthritis. Psoriatic arthritis, however, is usually not as severe, remissions are more frequent, and joint destruction is less frequent.

Bacterial Arthritis

Bacterial infections almost always cause an acute suppurative arthritis. The bacteria usually seed the joint during an episode of bacteremia; however, in neonates there is an increased incidence of contiguous spread from underlying epiphyseal osteomyelitis. The most common organisms are gonococcus, Staphylococcus, Streptococcus, Haemophilus influenzae, and gram-negative bacilli (E. coli, Salmonella, Pseudomonas, and others). H. influenzae arthritis predominates in children under 2 years of age, S. aureus is the main causative agent in older children and adults, and gonococcus is prevalent during late adolescence and young adulthood. Individuals with sickle cell disease are prone to infection with Salmonella at any age. These joint infections affect the sexes equally except for gonococcal arthritis, which is seen mainly in sexually active women. Predisposing conditions include immune deficiencies (congenital and acquired), debilitating illness, joint trauma, chronic arthritis of any cause, and intravenous drug abuse.

The classic presentation is the sudden development of an acutely painful and swollen infected joint that has a restricted range of motion. Systemic findings of fever, leukocytosis, and elevated sedimentation rate are common. In disseminated gonococcal infection the symptoms are more subacute. In 90% of nongonococcal cases, the infection involves only a single joint, usually the knee, followed in frequency by the hip, shoulder, elbow, wrist, and sternoclavicular joints. Axial articulations are more commonly involved in drug addicts. Prompt recognition and effective therapy prevent rapid joint destruction.

Tuberculous Arthritis

Tuberculous arthritis (Chapter 8) is a chronic progressive monoarticular disease that occurs in all age groups, especially adults. It usually develops as a complication of adjoining osteomyelitis or after hematogenous dissemination from a visceral (usually pulmonary) site of infection. Onset is insidious and causes gradual progressive pain. Systemic symptoms may or may not be present. Mycobacterial seeding of the joint induces the formation of confluent granulomas with central caseous necrosis. The affected synovium may grow as a pannus over the articular cartilage and erode the bone along the joint margins. Chronic disease results in severe destruction with fibrous ankylosis and obliteration of the joint space. The weight-bearing joints are usually affected, especially the hips, knees, and ankles in descending order of frequency.

Lyme Arthritis

As previously discussed (Chapter 8), Lyme arthritis is caused by infection with the spirochete Borrelia burgdorferi, which is transmitted by the ticks of the Ixodes ricinus complex. The initial infection of the skin is followed within several days or weeks by dissemination of the organism to other sites, especially the joints.

Approximately 60% to 80% of untreated individuals with Lyme disease develop joint symptoms within a few weeks to 2 years after the onset of the disease. The arthritis is the dominant feature of late disease; it tends to be remitting and migratory, and primarily involves large joints, especially the knees, shoulders, elbows, and ankles in descending order of frequency. Usually one or two joints are affected at a time, and the attacks last for a few weeks to months. Infected synovium exhibits a chronic papillary synovitis with synoviocyte hyperplasia, fibrin deposition, mononuclear cell infiltrates (especially CD4+ T cells), and onion-skin thickening of arterial walls. The morphology in severe cases can closely resemble that of rheumatoid arthritis. Silver stains may reveal small numbers of organisms in the vicinity of blood vessels in approximately 25% of cases. Chronic arthritis that is antibiotic refractory develops in approximately 10% of affected individuals and results from infection-induced autoimmunity. It is hypothesized that specific HLA-DR molecules bind an epitope of B. burgdorferi outer surface protein A, which initiates a T-cell reaction to this epitope. The T-cells may cross-react with an unknown self-antigen (an example of “molecular mimicry”). The joints in these patients have synovial pannus, which causes articular cartilage destruction and permanent deformities.⁵¹

Viral Arthritis

Arthritis can occur in the setting of a variety of viral infections, including alphavirus, parvovirus B19, rubella, Epstein-Barr virus, and hepatitis B and C virus. The clinical manifestations of the arthritis are variable and range from acute to subacute symptoms. It is unclear whether the joint symptoms are caused by direct infection of the joint by the virus, as seen in rubella and some alphavirus infections, or whether the viral infection generates an autoimmune reaction as seen in other forms of reactive or post-infectious arthritides.⁵² A variety of different rheumatic conditions, including reactive arthritis, psoriatic arthritis, and septic arthritis, have developed in individuals infected with HIV. The pathogenesis of some of these forms of HIV-associated chronic arthritis is probably autoimmune. The new effective antiretroviral therapies for HIV have ameliorated their severity.

Gout and Gouty Arthritis

Man is the only mammal to spontaneously develop hyperuricemia and gout, as only humans lack uricase, the enzyme responsible for the degradation of uric acid in other mammals. This, in combination with a high reabsorption rate of filtered urate, predisposes humans to hyperuricemia and gout, which is the common end point of a group of disorders that produce hyperuricemia.

Gout is marked by transient attacks of acute arthritis initiated by crystallization of urates within and about joints, leading eventually to chronic gouty arthritis and the appearance of tophi. Tophi represent large aggregates of urate crystals and the surrounding inflammatory reaction (see later). Most, but not all, individuals with chronic gout also develop urate nephropathy. Hyperuricemia (plasma urate level above 6.8 mg/dL) is necessary but not sufficient for the development of gout. More than 10% of the population of the Western hemisphere has hyperuricemia, and the prevalence is increasing; however, gout develops in fewer than 0.5% of these individuals. The various conditions producing hyperuricemia and gout (Table 26-7) are divided into those that produce primary gout (accounting for most idiopathic cases) and secondary gout (the cause of the hyperuricemia is known, and gout is not the main clinical expression of the disease). The role of hyperuricemia in the development of a variety of diseases, such as hypertension, chronic renal disease, cardiovascular disease, and the metabolic syndrome of hypertriglyceridemia, obesity, and insulin resistance, is controversial and remains a focus of investigation.

TABLE 26-7 Classification of Gout

Pathogenesis.

Uric acid is the end product of purine metabolism. Clinically, hyperuricemia develops from overproduction of urate in approximately 10% of cases (increased cell turnover—as in cancer, psoriasis, and during tumor lysis induced by chemotherapy) and reduced excretion in the remainder.

Plasma levels of uric acid are governed by a four-part renal transport system that involves glomerular filtration, reabsorption, secretion, and postsecretory reabsorption. Approximately 90% of the filtered urate is reabsorbed, and the urate transporter 1 gene (URAT1) has an important role in the reabsorption process. Decreased filtration and underexcretion of uric acid underlies most cases of primary gout. Two pathways are involved in purine synthesis: (1) a de novo pathway in which purines are synthesized from non-purine precursors and (2) a salvage pathway in which free purine bases derived from the breakdown of nucleic acids of endogenous or exogenous origin are recaptured (salvaged) (Fig. 26-46). The enzyme hypoxanthine guanine phosphoribosyl transferase (HGPRT) is involved in the salvage pathway. A deficiency of this enzyme leads to increased synthesis of purine nucleotides through the de novo pathway and hence increased production of uric acid. A complete lack of HGPRT occurs in the uncommon X-linked Lesch-Nyhan syndrome, seen only in males and characterized by hyperuricemia, severe neurologic deficits with mental retardation, self-mutilation, and in some cases gouty arthritis. Less severe deficiencies of the enzyme may also induce hyperuricemia and gouty arthritis with only mild neurologic deficits, but together these causes of gout are uncommon.

FIGURE 26-46 Purine metabolism. The conversion of PRPP to purine nucleotides is catalyzed by amido-PRT in the de novo pathway and by APRT and HGPRT in the salvage pathway. APRT, adenosine phosphoribosyltransferase; HGPRT, hypoxanthine-guanine phosporibosyltransferase; PRPP, phosphoribosyl pyrophosphate; PRT, phosphoribosyltransferase.

As stated earlier, hyperuricemia does not necessarily lead to gouty arthritis. Many factors contribute to the conversion of asymptomatic hyperuricemia into primary gout, including the following:

• Age of the individual and duration of the hyperuricemia. Gout rarely appears before 20 to 30 years of hyperuricemia.

• Genetic predisposition. In addition to the well-defined X-linked abnormalities of HGPRT, primary gout follows multifactorial inheritance and runs in families.

• Heavy alcohol consumption predisposes to attacks of gouty arthritis.

• Obesity increases the risk of asymptomatic gout.

• Certain drugs (e.g., thiazides) reduce excretion of urate and predispose to the development of gout.

• Lead toxicity increases the tendency to develop saturnine gout (Chapter 9).

Central to the pathogenesis of the arthritis is precipitation of monosodium urate (MSU) crystals into the joints (Fig. 26-47). The solubility of MSU in a joint is modulated by temperature (the lower the less soluble) and by the intra-articular concentration of urate and cations. Crystallization is dependent on the presence of nucleating agents such as insoluble collagen fibers, chondroitin sulfate, proteoglycans, cartilage fragments, and other crystals. Since synovial fluid is a poorer solvent for monosodium urate than plasma, urates in the joint fluid become supersaturated more easily, particularly in the peripheral joints (ankles and toes), where temperatures are as low as 20°C. With prolonged hyperuricemia, crystals and microtophi of urates develop in the synovium and in the joint cartilage. Some unknown event, possibly trauma, causes the release of crystals into the synovial fluid, which begins a cascade of events that initiates, intensifies, and sustains a powerful inflammatory response that is the hallmark of the acute attack. MSU crystals are phagocytosed by macrophages and through an incompletely understood mechanism activate the NALP3 inflammasome, a multiprotein complex that includes the protease caspase 1. The inflammasome-activated caspase 1, in turn, cleaves and activates several cytokines, most notably IL-1β and IL-18. IL-1β induces the expression of adhesion molecules and the synthesis of the neurophil chemokine CXCL8, which is essential for the localization of neutrophils at the site of acute inflammation. The neutrophils pour fuel on the fire by releasing toxic free radicals, leukotrienes (leukotriene B4), and lysosomal enzymes. Thus comes about an acute arthritis, which typically remits spontaneously in days to weeks.⁵³ A scheme of these events is shown in Figure 26-47.

FIGURE 26-47 Pathogenesis of acute gouty arthritis. LTB4, leukotriene B4.

Repeated attacks of acute arthritis lead eventually to chronic arthritis and the formation of tophi in the inflamed synovial membranes and periarticular tissue, as well as elsewhere. In time, severe damage to the cartilage develops and the function of the joints is compromised. It is not known why the chronic arthritis is asymptomatic for intervals of days to months, even though crystals are undoubtedly present in abundance in the joints.

Morphology. The distinctive morphologic changes in gout are (1) acute arthritis, (2) chronic tophaceous arthritis, (3) tophi in various sites, and sometimes (4) gouty nephropathy. Acute arthritis is characterized by a dense neutrophilic infiltrate that permeates the synovium and synovial fluid. The MSU crystals are frequently found in the cytoplasm of the neutrophils and are arranged in small clusters in the synovium. They are long, slender, and needle shaped, and are negatively birefringent. The synovium is edematous and congested, and also contains scattered lymphocytes, plasma cells, and macrophages. When the episode of crystallization abates and the crystals are resolubilized, the acute attack remits.

Chronic tophaceous arthritis evolves from the repetitive precipitation of urate crystals during acute attacks. The urates may heavily encrust the articular surfaces and form visible deposits in the synovium (Fig. 26-48). The synovium becomes hyperplastic, fibrotic, and thickened by inflammatory cells and forms a pannus that destroys the underlying cartilage leading to juxta-articular bone erosions. In severe cases, fibrous or bony ankylosis ensues, resulting in partial to complete loss of joint function.

FIGURE 26-48 Amputated great toe with white tophi involving the joint and soft tissues.

Tophi are the pathognomonic hallmark of gout. They are formed by large aggregations of urate crystals surrounded by an intense inflammatory reaction of macrophages, lymphocytes, and large foreign body giant cells, which may have completely or partially engulfed masses of crystals (Fig. 26-49). Tophi may appear in the articular cartilage of joints and in the periarticular ligaments, tendons, and soft tissues, including the olecranon and patellar bursae, Achilles tendons, and earlobes. Less frequently they may occur in the kidneys, nasal cartilages, skin of the fingertips, palms, soles, or elsewhere. Superficial tophi can ulcerate through the overlying skin.

FIGURE 26-49 Photomicrograph of a gouty tophus. An aggregate of dissolved urate crystals is surrounded by reactive fibroblasts, mononuclear inflammatory cells, and giant cells.

Gouty nephropathy (Chapter 20) is associated with the deposition of MSU crystals in the renal medullary interstitium, sometimes forming tophi, intratubular precipitations, or free uric acid crystals, and the production of uric acid renal stones. Secondary complications, such as pyelonephritis, may ensue, particularly when the urates induce some urinary obstruction.

Clinical Course.

The natural history of gout is said to have four stages: (1) asymptomatic hyperuricemia, (2) acute gouty arthritis, (3) intercritical gout, and (4) chronic tophaceous gout. Asymptomatic hyperuricemia appears around puberty in males and after menopause in females. After many years acute arthritis appears in the form of the sudden onset of excruciating joint pain associated with localized hyperemia, warmth, and exquisite tenderness. Yet constitutional symptoms are uncommon except possibly mild fever. Most first attacks are monoarticular; 50% occur in the first metatarsophalangeal joint. Eventually, about 90% of affected individuals experience acute attacks in the following locations (in descending order of frequency): insteps, ankles, heels, knees, wrists, fingers, and elbows. Untreated, acute gouty arthritis may last for hours to weeks, but gradually there is complete resolution and the patient enters an asymptomatic intercritical period. Although some patients never have another attack, most experience a second acute episode within months to a few years. In the absence of appropriate therapy, the attacks recur at shorter intervals and frequently become polyarticular. Eventually, over the span of years, disabling chronic tophaceous gout develops. On average about 12 years pass between the initial acute attack and the appearance of chronic tophaceous arthritis. At this stage, radiograms show characteristic juxta-articular bone erosion caused by osteoclastic bone resorption and loss of the joint space. Progression leads to severe crippling disease.

Cardiovascular disease including atherosclerosis and hypertension is common in individuals with gout. Renal manifestations sometimes appear in the form of renal colic associated with the passage of gravel and stones and may proceed to chronic gouty nephropathy. About 20% of those with chronic gout die of renal failure. The diagnosis of gout should not be delayed, because numerous drugs are available to abort or prevent acute attacks of arthritis and mobilize tophaceous deposits. Their use is important, because many aspects of the disease are related to the duration and severity of the hyperuricemia. Generally, gout does not materially shorten the life span, but it may impair the quality of life.

Calcium Pyrophosphate Crystal Deposition Disease (Pseudo-Gout)

Calcium pyrophosphate crystal deposition disease (CPPD), also known as pseudo-gout and chondrocalcinosis, is one of the more common disorders associated with intra-articular crystal formation. It usually occurs in individuals over 50 years of age and becomes more common with increasing age, rising to a prevalence of 30% to 60% in those 85 years or older. The sexes and races are equally affected. CPPD is divided into sporadic (idiopathic), hereditary, and secondary types. In the hereditary variant the crystals develop relatively early in life and are associated with severe osteoarthritis. The autosomal dominant form of the disease is caused by germline mutations in the ANKH gene, which encodes a transmembrane pyrophosphate transport channel.⁵⁴ The secondary form is associated with various disorders, including previous joint damage, hyperparathyroidism, hemochromatosis, hypomagnesemia, hypothyroidism, ochronosis, and diabetes. The basis for crystal formation is not known; altered activity of the matrix enzymes that produce and degrade pyrophosphate is suspected.

Morphology. The crystals first develop in the articular matrix, menisci, and intervertebral discs, and as the deposits enlarge they may rupture and seed the joint. Here, they are phagocytosed by macrophages, in which they activate the NALP3 inflammasome, eliciting a series of pro-inflammatory events similar or identical to those induced by urate crystals (Fig. 26-47). Neutrophils recruited by inflammatory mediators are thought to produce damage through the release of reactive oxygen species, catabolic enzymes, and cytokines, calling forth the more chronic reactions associated with macrophages and fibrosis. The crystals form chalky white friable deposits, which are seen histologically in stained preparations as oval blue-purple aggregates. Individual crystals are generally 0.5 to 5 μm in greatest dimension, are weakly birefringent, and have geometric shapes (Fig. 26-50). Rarely the crystals are deposited in masslike aggregates simulating tophi.

FIGURE 26-50 Smear preparation of calcium pyrophosphate crystals.

CPPD is frequently asymptomatic; however, it also can produce acute, subacute, or chronic arthritis that can be confused with osteoarthritis or rheumatoid arthritis. The joint involvement may last from several days to weeks and may be monoarticular or polyarticular; the knees, followed by the wrists, elbows, shoulders, and ankles, are most commonly affected. Ultimately, approximately 50% of affected individuals experience significant joint damage. Therapy is supportive. There is no known treatment that prevents or retards crystal formation.

Nodular Fasciitis

Nodular fasciitis, also known as infiltrative or pseudosarcomatous fasciitis, is the most common of the reactive pseudosarcomas. It most often occurs in adults on the volar aspect of the forearm, followed in order of frequency by the chest and back. Affected individuals typically present with a several-week history of a solitary, rapidly growing, and sometimes painful mass. Preceding trauma is reported in only 10% to 15% of cases.

Morphology. Nodular fasciitis arises in the deep dermis, subcutis, or muscle. Grossly the lesion is several centimeters in greatest dimension, is nodular in configuration, and has poorly defined margins. The lesion is richly cellular and contains plump, immature-appearing fibroblasts and myofibroblasts arranged randomly or in short intersecting fascicles (Fig. 26-54). The cells vary in size and shape (spindle to stellate) and have conspicuous nucleoli; mitotic figures are abundant. Frequently the stroma is myxoid and contains lymphocytes and extravasated red blood cells. The histologic differential is extensive, but important lesions that must be excluded are fibromatosis and spindle cell sarcomas. Nodular fasciitis rarely recurs after excision.

FIGURE 26-54 Nodular fasciitis with plump, randomly oriented spindle cells surrounded by myxoid stroma. Note the mitotic activity and extravasated red blood cells.

Myositis Ossificans

Myositis ossificans is distinguished from the other reactive fibroblastic proliferations by the presence of metaplastic bone. It usually develops in athletic adolescents and young adults and follows an episode of trauma in more than 50% of cases. The lesion typically arises in the musculature of the proximal extremities. The clinical findings are related to its stage of development; in the early phase the involved area is swollen and painful, and during the subsequent several weeks it becomes more circumscribed and firm. Eventually, it evolves into a painless, hard, well-demarcated mass.

The radiographic findings parallel the morphologic progression. Initially, X-rays may show only soft-tissue fullness, but at about 3 weeks patchy flocculent radiodensities form in the periphery. The radiodensities become more extensive with time and slowly encroach on the radiolucent center (Fig. 26-55). Myositis ossificans must be distinguished from extraskeletal osteosarcoma. The latter usually occurs in elderly patients, the proliferating cells are cytologically malignant, and the tumor lacks the zonation of myositis ossificans. Simple excision of myositis ossificans is usually curative.

FIGURE 26-55 Peripherally mineralized myositis ossificans (arrows) involving the posterior thigh.

Superficial Fibromatosis (Palmar, Plantar, and Penile Fibromatoses)

Palmar, plantar, and penile fibromatoses, lesions that are more bothersome than serious, constitute a small group of superficial fibromatoses. They are characterized by nodular or poorly defined broad fascicles of fibroblasts and myofibroblasts surrounded by abundant dense collagen. The molecular mechanisms underlying superficial fibromatoses are unknown, but they are different from their deep-seated counterparts.

In the palmar variant (Dupuytren contracture) there is irregular or nodular thickening of the palmar fascia either unilaterally or bilaterally (50%). Over a span of years, attachment tothe overlying skin causes puckering and dimpling. At the same time a slowly progressive flexion contracture develops that mainly affects the fourth and fifth fingers of the hand. Essentially similar changes are seen with plantar fibromatosis except that flexion contractures are uncommon and bilateral involvement is infrequent. In penile fibromatosis (Peyronie disease) a palpable induration or mass appears, usually on the dorsolateral aspect of the penis. Eventually, it may cause abnormal curvature of the shaft, constriction of the urethra, or both.

All forms of superficial fibromatosis affect males more frequently than females. In about 20% to 25% of cases, the palmar and plantar fibromatoses stabilize and do not progress, in some instances resolving spontaneously. Some recur after excision, particularly the plantar variant.

Deep-Seated Fibromatosis (Desmoid Tumors)

Deep-seated fibromatoses lie in a gray zone between benign fibrous tumors and low-grade fibrosarcomas. On the one hand, they often present as large, infiltrative masses that frequently recur after incomplete excision; on the other, they are composed of banal well-differentiated fibroblasts that do not metastasize. They may occur at any age but are most frequent in the teens to 30s. Deep-seated fibromatosis is divided into extra-abdominal, abdominal, and intra-abdominal types, but all have similar gross and microscopic features. Extra-abdominal fibromatosis occurs in men and women with equal frequency and arises principally in the musculature of the shoulder, chest wall, back, and thigh. Abdominal fibromatosis generally arises in the musculoaponeurotic structures of the anterior abdominal wall in women during or after pregnancy. Intra-abdominal fibromatosis tends to occur in the mesentery or pelvic walls, often in individuals having familial adenomatous polyposis (Gardner syndrome) (Chapter 17). Mutations in the APC or β-catenin genes are present in the majority of these tumors (whether or not the affected individuals have underlying Gardner syndrome) and have an important role in their genesis.

Morphology. These tumors occur as gray-white, firm, poorly demarcated masses varying from 1 to 15 cm in greatest diameter. They are rubbery and tough, and infiltrate surrounding structures. Histologically deep-seated fibromatosis is composed of plump banal fibroblasts arranged in broad sweeping fascicles that infiltrate the adjacent tissue (Fig. 26-56). Mitoses may be frequent. Regenerative muscle cells when trapped within these lesions may take on the appearance of multinucleated giant cells.

FIGURE 26-56 Fibromatosis infiltrating between skeletal muscle cells.

In addition to possibly being disfiguring or disabling, deep-seated fibromatosis is occasionally painful. Although curable by adequate excision, these lesions frequently recur locally and persistently when incompletely removed. Some tumors have responded to treatment with tamoxifen, and in other cases chemotherapy or irradiation has been effective. The rare reports of metastasis of fibromatosis must be interpreted as misdiagnosis of fibrosarcoma.