JOURNAL OF BONE AND MINERAL RESEARCH Volume 6, Supplement 2, 1991 Mary Ann Liebert, Inc., Publishers
Interpretation of Bone Mass Determinations as They Relate to Fracture: Implications for Asymptomatic Primary Hyperparathyroidism MUNRO PEACOCK
ABSTRACT Determination of bone mass is currently the most clinically useful measurement of bone strength and of fracture risk. Interpretation of bone mass determination as it relates to fracture has been developed largely from studies of age-related bone loss. A decrease in bone mass and an increase in fracture incidence with aging are universal phenomena that are causally related by virtue of the major contribution bone mass makes to skeletal strength. Over 70% of the skeleton's strength to resist fracture resides in its mineral content in vitro. Clinically, the relationship between trauma and fracture is complex, and in the general population, fracture appears as a random event occurring more frequently as bone mass decreases. In the individual, measurement of bone mass in relation to the range of bone mass and the fracture incidence of the reference population provides an estimate of the risk of sustaining a fracture in the future. In primary hyperparathyroidism, interpretation of a bone mass determination must take into account the effect of the disease activity on the skeleton against the background of universal age-related changes in bone mass and fracture incidence. This general relationship is likely to be altered by at least three unique effects that parathyroid hormone may have on the skeleton: (1)parathyroid hormone has a differential effect on cortical and cancellous bone; (2) it has a biphasic effect on bone that is concentration dependent; and (3) it alters bone quality and architecture. In asymptomatic patients with primary hyperparathyroidism who are considered for long-term medical follow-up, bone mass determination and fracture incidence must be part of the essential monitoring process both to provide a best estimate of fracture risk for the subject and to provide prospective data to refine the estimate of fracture risk for the disease itself.
INTRODUCTION are currently the most clinically useful measurements of bone strength and thus of the risk of sustaining a fracture of the skeleton. Fracture is a major clinical problem, particularly in the elderly, and occurs when the skeleton is insufficiently strong to absorb the mechanical forces to which it is subjected. It results either from excessive trauma or from excessive weakness of the bone secondary to a change in the quality of the osteoid, the architectural integrity of cancellous and cortical bone, and/or the mass of bone. Mineral content is the major determinant of bone strength in vitro.
D
ETERMINATIONS OF BONE MASS
Clinically, the relationship between the trauma causing fracture and bone mass is more complex, and in the general population fracture appears as a random event whose frequency increases as bone mass decreases. Bone mass changes markedly with age.") I t has a wide normal range that is set by a variety of factors in addition or related to age. Differences in bone mass and its rate of loss in themselves would have little clinical importance but for the fact that as bone mass decreases there is a rise in incidence of fractures.'2) This inverse relationship is universal, and low bone mass is associated with about 1.3 million fractures each year in the United States.13) In severe, symptomatic primary hyperparathyroidism,
Department of Medicine, Indiana University School of Medicine, Indianapolis, Indiana
s77
PEACOCK
S78
fractures are a recognized feature of the disease and can occur anywhere in the ~keleton.'~) In mild, asymptomatic primary hyperparathyroidism, an unresolved question in management is whether there is a parathyroid-dependent increase in fracture incidence, and if so, whether it is due to an abnormality in bone mass, bone architecture, and/or bone quality. It might be assumed that asymptomatic primary hyperparathyroidism is merely a milder form of primary hyperparathyroidism and fracture is less frequent or is a later event. However, there is evidence suggesting that PTH has dual actions on the skeleton, perhaps depending on its serum concentration, such that high concentrations are catabolic, particularly on cortical bone, whereas marginally elevated concentrations, as occur in asymptomatic primary hyperparathyroidism, are anabolic, particularly on cancellous bone.(5' Indeed, with the anabolic effect in mind, parathyroid hormone (PTH) has been studied as potential treatment for osteoporosis.(6) The interpretation of bone mass determinations as they relate to fracture have evolved mainly from studies of agerelated bone loss. In patients with asymptomatic primary hyperparathyroidism who are managed by medical followup, regular bone mass determination and fracture incidence need to be a part of the follow-up process. It remains to be established, however, whether the interpretation of bone mass determinations, as developed in age-related bone loss, are completely transferable to a disease like primary hyperparathyroidism that has unique actions on bone.
Severe trauma results in fracture. In the elderly and in subjects with low bone mass, however, the majority of fractures arise from minimal trauma, often as a result of simple falls. The incidence of falls is positively related to age,(1s)and the interaction of the frequency of falls and low bone mass is an important determinant of fracture incidence.
BONE MASS
The bone mass of an individual is set by the amount of peak bone mass laid down, by the age of onset of bone loss, and by the rate of bone loss. A wide variety of factors, in addition to age, affect all or some of these three processes (Table 2) and contribute to the strong inverse relationship between bone mass and age in the general population. During the first two decades of life, bone mass increases with g r o ~ t h . ( ' ~ .In ' ~ the ) third decade, the skeleton continues to consolidate and bone mass increases, although at a much reduced ~ a t e . 1 ' ~Bone ) mass reaches its maximum and remains relatively stable in the fourth decade of life, with women having less bone than men''.') and inheritance(18)and ethni~ity"~) being important determinants. From the fifth decade onward, age-related bone loss is universal and on average amounts to l%/year.(z) It affects cancellous and cortical bone and is accompanied by changes in skeletal architecture. Tubular bones expand in width,('6) cortical thickness decreases,(16)and trabeculae thin and may totally disappear.(12.L3) Mechanical forces are essential in maintaining skeletal mass. However, they are poorly understood in biochemical FRACTURE terms,(20)and relationships between different types of activity and bone mass remain to be established.(2')Bone Bone fractures when force is applied to the skeleton that growth and remodeling are closely regulated by a wide it cannot totally absorb, because either the force is exces- range of hormones (Table 2). Overt diseases of over- and sive or the bone is weak. Bone strength in vivo is difficult undersecretion of these hormones lead to well-recognized to measure directly, although techniques are currently abnormalities in the skeleton. However, the effect on bone under development that may prove useful.(7)In vitro stud- of more subtle changes in secretion, particularly with ies indicate that over 70% of bone strength is due to its aging, are less clear. Perhaps the best appreciated of these ~ ~ ~ ~ ~in) growth hormone and mineral content (Table 1). (8-'1) Architectural changes also is ovarian f a i l u ~ e . (Changes ~~' androplay a r ~ l e , ( ~although ~ , ~ ~ many ) are themselves related to the ~ o m a t o m e d i n s , ( ~t e~ s) t o s t e r ~ n e , ~adrenal gens,(26)parathyroid hormone,(27)and vitamin D(z81have loss in bone mineral content. Over the age of 50 there is marked age-related increase been implicated as factors in age-related bone loss. Calaffect the rate of age-rein most fractures. (I4) Three fractures, however, predomi- cium intaketz9)and nate: forearm fracture increasing in incidence in the fifth lated bone loss, and such dietary constituents as pro' ) also play a role either directly or indirectly decade, vertebral fracture increasing in the sixth decade, t e i r ~ ' ~may and hip fracture in the seventh decade. All three fractures through the ability to adapt to dietary calcium changes. A * - ~affect ~ ) bone are strongly age-related, with women having an overall variety of factors related to l i f e - ~ t y l e ~ ~also mass. fracture incidence two to five times that of men. (yo R ') EXPLAINED BY BONEMINERAL TABLE1. VARIATIONIN BONESTRENGTH CONTENT IN V I T R O ~
Femur calcar(B) 85 yo
Femur neck(9'
Lumbar vertebrae"')
Distal and proximal radius'"'
77%
74%
62qo
aR' = index of determination.
BONE MASS A N D FRACTURE IN HYPERPARATHYROIDISM
s79
ety of studies [see review by Ross et a1.(40)]that measurement of bone mass predicts fracture risk.l3' From crosssectional studies, estimates of the risk of fracture due to Age low bone mass in subjects who have already fractured are Gender obtained. Prospective studies are now available, supplying Heredity much stronger evidence that a single bone mass measureEthnicity ment identifies those at risk of future fracture (Table 3). Mechanical force Because of the high correlation between bone mass at Disease different skeletal sites, fracture risk can be estimated from Life-style Alcohol bone mass measured at various sites. It is likely that bone Tobacco mass measured at the site of future fracture provides the Medications best prediction of risk. However, it should be noted that Diet fracture risk may not be extrapolated from one population Calcium or from one age group to to Protein Risk factors for fracture may be combined with bone Hormones mass measurement t o improve prediction. In a prospective Estrogen study on hip fracture by Porter and his colleagues,(46)bone Androgen mass, assessed by ultrasound attenuation at the 0s calcis, Calcitriol/calcidiol gave a relative fracture risk of 2.8, whereas combined with Parathyroid hormone Growth hormone a score of cognizance, the relative risk increased to 8.5. This indicates that the more risk factors for fracture that can be identified in a population, the higher the prediction becomes. The increase in bone mass induced by a treatment can be used to predict its cost On the other hand, such treatments as fluoride increase bone mass withFRACTURE AND BONE MASS: out decreasing fracture and the quantity and RISK-FACTOR MODEL quality of the bone laid down are both important. Evaluation of preventative treatment requires measurements of Radiography and photon absorptiometry quantify bone bone mass and fracture incidence. mass both morphometrically and densitometrically.~'6.35) These measurements, however, d o not separate patients with fracture from the normal age- and sex-matched population. This overlap has led some to believe that the amount of bone is at most a minor component in the pathogenesis of fracture and that measurement of bone mass of PRIMARY HYPERPARATHYROIDISM little value in the investigation and management of osteoDisease activity and classification porotic fracture. (36,371 Newton-John and Morgan,'*' reviewing all published The spectrum of activity of disease in primary hyperdata on both fracture incidence and bone mass measure- parathyroidism is wide. As judged by the increase in serum ments, concluded that bone loss with age is universal, that calcium level at presentation, disease activity ranges from the amount of bone present determines the risk of frac- 1% of normal to over 60%. With the widespread use of ture, and that the increase in frequency of fracture with multichannel biochemical screening tests that include age is largely determined by the normal loss of bone with serum calcium, the prevalence of primary hyperparathyage. They defined a fracture threshold based on the range roid patients having minimal disease activity and few or no of peak bone mass of the healthy adult population and signs and symptoms has greatly increased. ( 4 9 1 Such patients showed that the more bone mass decreased below this have been classified as asymptomatic primary hyperparavalue, the greater became the risk of fracture. The study thyroidism to distinguish them from patients who require by Smith and his c o l l e a g ~ e sof ~ ~osteoporotic ~) and normal parathyroidectomy to treat stone formation, radiologic women strongly supported this model and, importantly, bone disease, and/or hypercalcemic Because showed prospectively that future fracture could be pre- of the difficulty of finding a minimally enlarged gland at dicted from measurement of bone mass. This risk-factor parathyroidectomy and the clinical experience that disease model suggests that fracture is a random event whose inci- activity may remain unchanged for many years, there has dence in a population is determined by the amount of bone been a trend in many centers to follow asymptomatic priin the skeleton, and by using the variables, peak bone mary hyperparathyroid patients medically and postpone mass, age of onset of bone loss, and rate of bone loss with surgical treatment unless symptoms develop. Such an apage, the incidence of fracture in a population can be accu- proach, however, ignores the possibility that there may be a PTH-dependent loss of bone and an increase in incidence rately Substantial evidence has now accumulated from a vari- of fracture several decades after diagnosis. Neither the criTABLE2. DETERMINANTS OF BONEMASS
S80
PEACOCK TABLE3 . PREDICTION
Reference Population Age Fracture site Odds ratio
42 1078 Free-living 43-90 All 2.8
OF
FUTUREFRACTURE IN WOMEN FROM A SINGLE MEASUREMENT OF BONEMASS AT THE PROXIMAL RADIUS~
43 1076 White Free-living 30-90 All except asymptomatic spine I .6
41 386 White Free-living 40-80 All except spine 6.7
41 135 White Retirement home 50-90 All except spine 2.6
44 9707 White Free-living 65-90 Hip 2.3
aOdds ratio calculated on a 2SD difference in bone mass.""'
teria for diagnosing asymptomatic hyperparathyroidism nor those for advising parathyroidectomy currently take into account a measurement of bone mass.'5o)
Action of PTH on bone Although PTH has been recognized for many years as a major regulator of skeletal metabolism, many of its actions on bone are still poorly understood. In severe hyperparathyroidism the effects on the skeleton are well recognized and cause changes in bone mass, architecture, and quality. Bone resorption and formation are greatly increased. Focal subcortical erosions and cystic lesions due to osteoclast proliferation are often the site of fracture. Stimulation of fibrosis in the marrow may be marked. Clinically, PTH appears to stimulate bone turnover by increasing the remodeling space. Once a steady state is achieved, a set quantum of bone is The skeleton is permanently in negative balance by this amount even though the net rate of bone loss may return to normal. Only in those subjects whose hyperparathyroidism progresses is the rate of bone loss abnormal. On the other hand, in short-term animal studies, parenteral PTH increases the amount of cancellous bone without reducing cortical bone. The overall affect being a net increase in bone mass. Similar effects on cancellous bone are seen in humans with primary osteoporosis given daily injections of PTH in amounts insufficient to cause h y p e r ~ a l c e m i a . ' ~In. ~these ~ ' studies, however, there was a reduction in cortical bone and also a marked variability in the responses between subjects. The level of P T H activity needed to induce anabolic or catabolic effects on bone and the variability between subjects are important questions that await an answer. Nevertheless, the speculation that mild hyperparathyroidism may equate to the level of PTH activity that is anabolic suggests that patients with primary hyperparathyroidism may not form a continuous spectrum for either bone mass or fracture.
Bone mass and fracture Studies addressing the question of cortical and cancellous bone loss in asymptomatic primary hyperparathyroidism are few. In 52 patients with asymptomatic primary hy-
perparathyroidism who had iliac crest biopsies and mineral density measured at the radius, spine, and hip, a deficit in cortical but not in cancellous bone was Indeed, the latter tended to be higher than normal, suggesting a protective effect of P T H at skeletal sites with a major cancellous component. The same findings applied whether the mild hyperparathyroidism was or was not associated with stone formation.(54' In a study involving 174 black and white men and women with asymptomatic primary hyperparathyroidism (serum calcium 5 12 mg%), vertebral fractures occurred but the prevalence assessed from radiographs was normal.'ss) There was also no increase in overall fracture prevalence, although bone mass in the forearm was reduced. On the other hand, in 174 women with primary hyperparathyroidism there was an increased prevalence of fracture at the wrist, hip, and spine (Table 4 ) . ( 5 1 1 In these patients with fracture, 72% had a serum calcium I 12 mg% and in 57% this was 5 11.25 mg%. Both cortical bone (metacarpal cortical width) and cancellous bone (Singh grade) were reduced, and the cortical deficit corrected for age was more marked in the elderly patients. The fracture prevalence and bone deficit were highest in those with the lowest absorption of calcium. The rate of cortical bone loss in patients with mild primary hyperparathyroidism is only marginally increased, although parathyroidectomy results in a significant decrease in the rate of loss in the short term.(51)In the long term, however, there is incomplete recovery of the bone defiThese differences in fracture incidence may reflect differences between centers in selection criteria and in methods of measuring fracture occurrence. O n the other hand, they may reflect true geographic variation in fracture prevalence and indicate the need for more studies. TABLE 4. NUMBER OF OBSERVED A N D EXPECTED FRACTURES IN 174 WOMEN WITH PRIMARY HYPERPARATHYROIDISM Skeletal site
Observed
Expected
P
26 23 12 5 4 4
3.9 9.0 1.8 < 0.2 < 0.2 < 0.2
< 0.001 < 0.001 < 0.001 < 0.001
Spine Wrist Femur Wrist and spine Wrist and femur Spine and femur Source: From Ref. 5 1 .
Interpretation of Bone Mass Determinations as They Relate to Fracture: Implications for Asymptomatic Primary Hyperparathyroidism MUNRO PEACOCK
ABSTRACT Determination of bone mass is currently the most clinically useful measurement of bone strength and of fracture risk. Interpretation of bone mass determination as it relates to fracture has been developed largely from studies of age-related bone loss. A decrease in bone mass and an increase in fracture incidence with aging are universal phenomena that are causally related by virtue of the major contribution bone mass makes to skeletal strength. Over 70% of the skeleton's strength to resist fracture resides in its mineral content in vitro. Clinically, the relationship between trauma and fracture is complex, and in the general population, fracture appears as a random event occurring more frequently as bone mass decreases. In the individual, measurement of bone mass in relation to the range of bone mass and the fracture incidence of the reference population provides an estimate of the risk of sustaining a fracture in the future. In primary hyperparathyroidism, interpretation of a bone mass determination must take into account the effect of the disease activity on the skeleton against the background of universal age-related changes in bone mass and fracture incidence. This general relationship is likely to be altered by at least three unique effects that parathyroid hormone may have on the skeleton: (1)parathyroid hormone has a differential effect on cortical and cancellous bone; (2) it has a biphasic effect on bone that is concentration dependent; and (3) it alters bone quality and architecture. In asymptomatic patients with primary hyperparathyroidism who are considered for long-term medical follow-up, bone mass determination and fracture incidence must be part of the essential monitoring process both to provide a best estimate of fracture risk for the subject and to provide prospective data to refine the estimate of fracture risk for the disease itself.
INTRODUCTION are currently the most clinically useful measurements of bone strength and thus of the risk of sustaining a fracture of the skeleton. Fracture is a major clinical problem, particularly in the elderly, and occurs when the skeleton is insufficiently strong to absorb the mechanical forces to which it is subjected. It results either from excessive trauma or from excessive weakness of the bone secondary to a change in the quality of the osteoid, the architectural integrity of cancellous and cortical bone, and/or the mass of bone. Mineral content is the major determinant of bone strength in vitro.
D
ETERMINATIONS OF BONE MASS
Clinically, the relationship between the trauma causing fracture and bone mass is more complex, and in the general population fracture appears as a random event whose frequency increases as bone mass decreases. Bone mass changes markedly with age.") I t has a wide normal range that is set by a variety of factors in addition or related to age. Differences in bone mass and its rate of loss in themselves would have little clinical importance but for the fact that as bone mass decreases there is a rise in incidence of fractures.'2) This inverse relationship is universal, and low bone mass is associated with about 1.3 million fractures each year in the United States.13) In severe, symptomatic primary hyperparathyroidism,
Department of Medicine, Indiana University School of Medicine, Indianapolis, Indiana
s77
PEACOCK
S78
fractures are a recognized feature of the disease and can occur anywhere in the ~keleton.'~) In mild, asymptomatic primary hyperparathyroidism, an unresolved question in management is whether there is a parathyroid-dependent increase in fracture incidence, and if so, whether it is due to an abnormality in bone mass, bone architecture, and/or bone quality. It might be assumed that asymptomatic primary hyperparathyroidism is merely a milder form of primary hyperparathyroidism and fracture is less frequent or is a later event. However, there is evidence suggesting that PTH has dual actions on the skeleton, perhaps depending on its serum concentration, such that high concentrations are catabolic, particularly on cortical bone, whereas marginally elevated concentrations, as occur in asymptomatic primary hyperparathyroidism, are anabolic, particularly on cancellous bone.(5' Indeed, with the anabolic effect in mind, parathyroid hormone (PTH) has been studied as potential treatment for osteoporosis.(6) The interpretation of bone mass determinations as they relate to fracture have evolved mainly from studies of agerelated bone loss. In patients with asymptomatic primary hyperparathyroidism who are managed by medical followup, regular bone mass determination and fracture incidence need to be a part of the follow-up process. It remains to be established, however, whether the interpretation of bone mass determinations, as developed in age-related bone loss, are completely transferable to a disease like primary hyperparathyroidism that has unique actions on bone.
Severe trauma results in fracture. In the elderly and in subjects with low bone mass, however, the majority of fractures arise from minimal trauma, often as a result of simple falls. The incidence of falls is positively related to age,(1s)and the interaction of the frequency of falls and low bone mass is an important determinant of fracture incidence.
BONE MASS
The bone mass of an individual is set by the amount of peak bone mass laid down, by the age of onset of bone loss, and by the rate of bone loss. A wide variety of factors, in addition to age, affect all or some of these three processes (Table 2) and contribute to the strong inverse relationship between bone mass and age in the general population. During the first two decades of life, bone mass increases with g r o ~ t h . ( ' ~ .In ' ~ the ) third decade, the skeleton continues to consolidate and bone mass increases, although at a much reduced ~ a t e . 1 ' ~Bone ) mass reaches its maximum and remains relatively stable in the fourth decade of life, with women having less bone than men''.') and inheritance(18)and ethni~ity"~) being important determinants. From the fifth decade onward, age-related bone loss is universal and on average amounts to l%/year.(z) It affects cancellous and cortical bone and is accompanied by changes in skeletal architecture. Tubular bones expand in width,('6) cortical thickness decreases,(16)and trabeculae thin and may totally disappear.(12.L3) Mechanical forces are essential in maintaining skeletal mass. However, they are poorly understood in biochemical FRACTURE terms,(20)and relationships between different types of activity and bone mass remain to be established.(2')Bone Bone fractures when force is applied to the skeleton that growth and remodeling are closely regulated by a wide it cannot totally absorb, because either the force is exces- range of hormones (Table 2). Overt diseases of over- and sive or the bone is weak. Bone strength in vivo is difficult undersecretion of these hormones lead to well-recognized to measure directly, although techniques are currently abnormalities in the skeleton. However, the effect on bone under development that may prove useful.(7)In vitro stud- of more subtle changes in secretion, particularly with ies indicate that over 70% of bone strength is due to its aging, are less clear. Perhaps the best appreciated of these ~ ~ ~ ~ ~in) growth hormone and mineral content (Table 1). (8-'1) Architectural changes also is ovarian f a i l u ~ e . (Changes ~~' androplay a r ~ l e , ( ~although ~ , ~ ~ many ) are themselves related to the ~ o m a t o m e d i n s , ( ~t e~ s) t o s t e r ~ n e , ~adrenal gens,(26)parathyroid hormone,(27)and vitamin D(z81have loss in bone mineral content. Over the age of 50 there is marked age-related increase been implicated as factors in age-related bone loss. Calaffect the rate of age-rein most fractures. (I4) Three fractures, however, predomi- cium intaketz9)and nate: forearm fracture increasing in incidence in the fifth lated bone loss, and such dietary constituents as pro' ) also play a role either directly or indirectly decade, vertebral fracture increasing in the sixth decade, t e i r ~ ' ~may and hip fracture in the seventh decade. All three fractures through the ability to adapt to dietary calcium changes. A * - ~affect ~ ) bone are strongly age-related, with women having an overall variety of factors related to l i f e - ~ t y l e ~ ~also mass. fracture incidence two to five times that of men. (yo R ') EXPLAINED BY BONEMINERAL TABLE1. VARIATIONIN BONESTRENGTH CONTENT IN V I T R O ~
Femur calcar(B) 85 yo
Femur neck(9'
Lumbar vertebrae"')
Distal and proximal radius'"'
77%
74%
62qo
aR' = index of determination.
BONE MASS A N D FRACTURE IN HYPERPARATHYROIDISM
s79
ety of studies [see review by Ross et a1.(40)]that measurement of bone mass predicts fracture risk.l3' From crosssectional studies, estimates of the risk of fracture due to Age low bone mass in subjects who have already fractured are Gender obtained. Prospective studies are now available, supplying Heredity much stronger evidence that a single bone mass measureEthnicity ment identifies those at risk of future fracture (Table 3). Mechanical force Because of the high correlation between bone mass at Disease different skeletal sites, fracture risk can be estimated from Life-style Alcohol bone mass measured at various sites. It is likely that bone Tobacco mass measured at the site of future fracture provides the Medications best prediction of risk. However, it should be noted that Diet fracture risk may not be extrapolated from one population Calcium or from one age group to to Protein Risk factors for fracture may be combined with bone Hormones mass measurement t o improve prediction. In a prospective Estrogen study on hip fracture by Porter and his colleagues,(46)bone Androgen mass, assessed by ultrasound attenuation at the 0s calcis, Calcitriol/calcidiol gave a relative fracture risk of 2.8, whereas combined with Parathyroid hormone Growth hormone a score of cognizance, the relative risk increased to 8.5. This indicates that the more risk factors for fracture that can be identified in a population, the higher the prediction becomes. The increase in bone mass induced by a treatment can be used to predict its cost On the other hand, such treatments as fluoride increase bone mass withFRACTURE AND BONE MASS: out decreasing fracture and the quantity and RISK-FACTOR MODEL quality of the bone laid down are both important. Evaluation of preventative treatment requires measurements of Radiography and photon absorptiometry quantify bone bone mass and fracture incidence. mass both morphometrically and densitometrically.~'6.35) These measurements, however, d o not separate patients with fracture from the normal age- and sex-matched population. This overlap has led some to believe that the amount of bone is at most a minor component in the pathogenesis of fracture and that measurement of bone mass of PRIMARY HYPERPARATHYROIDISM little value in the investigation and management of osteoDisease activity and classification porotic fracture. (36,371 Newton-John and Morgan,'*' reviewing all published The spectrum of activity of disease in primary hyperdata on both fracture incidence and bone mass measure- parathyroidism is wide. As judged by the increase in serum ments, concluded that bone loss with age is universal, that calcium level at presentation, disease activity ranges from the amount of bone present determines the risk of frac- 1% of normal to over 60%. With the widespread use of ture, and that the increase in frequency of fracture with multichannel biochemical screening tests that include age is largely determined by the normal loss of bone with serum calcium, the prevalence of primary hyperparathyage. They defined a fracture threshold based on the range roid patients having minimal disease activity and few or no of peak bone mass of the healthy adult population and signs and symptoms has greatly increased. ( 4 9 1 Such patients showed that the more bone mass decreased below this have been classified as asymptomatic primary hyperparavalue, the greater became the risk of fracture. The study thyroidism to distinguish them from patients who require by Smith and his c o l l e a g ~ e sof ~ ~osteoporotic ~) and normal parathyroidectomy to treat stone formation, radiologic women strongly supported this model and, importantly, bone disease, and/or hypercalcemic Because showed prospectively that future fracture could be pre- of the difficulty of finding a minimally enlarged gland at dicted from measurement of bone mass. This risk-factor parathyroidectomy and the clinical experience that disease model suggests that fracture is a random event whose inci- activity may remain unchanged for many years, there has dence in a population is determined by the amount of bone been a trend in many centers to follow asymptomatic priin the skeleton, and by using the variables, peak bone mary hyperparathyroid patients medically and postpone mass, age of onset of bone loss, and rate of bone loss with surgical treatment unless symptoms develop. Such an apage, the incidence of fracture in a population can be accu- proach, however, ignores the possibility that there may be a PTH-dependent loss of bone and an increase in incidence rately Substantial evidence has now accumulated from a vari- of fracture several decades after diagnosis. Neither the criTABLE2. DETERMINANTS OF BONEMASS
S80
PEACOCK TABLE3 . PREDICTION
Reference Population Age Fracture site Odds ratio
42 1078 Free-living 43-90 All 2.8
OF
FUTUREFRACTURE IN WOMEN FROM A SINGLE MEASUREMENT OF BONEMASS AT THE PROXIMAL RADIUS~
43 1076 White Free-living 30-90 All except asymptomatic spine I .6
41 386 White Free-living 40-80 All except spine 6.7
41 135 White Retirement home 50-90 All except spine 2.6
44 9707 White Free-living 65-90 Hip 2.3
aOdds ratio calculated on a 2SD difference in bone mass.""'
teria for diagnosing asymptomatic hyperparathyroidism nor those for advising parathyroidectomy currently take into account a measurement of bone mass.'5o)
Action of PTH on bone Although PTH has been recognized for many years as a major regulator of skeletal metabolism, many of its actions on bone are still poorly understood. In severe hyperparathyroidism the effects on the skeleton are well recognized and cause changes in bone mass, architecture, and quality. Bone resorption and formation are greatly increased. Focal subcortical erosions and cystic lesions due to osteoclast proliferation are often the site of fracture. Stimulation of fibrosis in the marrow may be marked. Clinically, PTH appears to stimulate bone turnover by increasing the remodeling space. Once a steady state is achieved, a set quantum of bone is The skeleton is permanently in negative balance by this amount even though the net rate of bone loss may return to normal. Only in those subjects whose hyperparathyroidism progresses is the rate of bone loss abnormal. On the other hand, in short-term animal studies, parenteral PTH increases the amount of cancellous bone without reducing cortical bone. The overall affect being a net increase in bone mass. Similar effects on cancellous bone are seen in humans with primary osteoporosis given daily injections of PTH in amounts insufficient to cause h y p e r ~ a l c e m i a . ' ~In. ~these ~ ' studies, however, there was a reduction in cortical bone and also a marked variability in the responses between subjects. The level of P T H activity needed to induce anabolic or catabolic effects on bone and the variability between subjects are important questions that await an answer. Nevertheless, the speculation that mild hyperparathyroidism may equate to the level of PTH activity that is anabolic suggests that patients with primary hyperparathyroidism may not form a continuous spectrum for either bone mass or fracture.
Bone mass and fracture Studies addressing the question of cortical and cancellous bone loss in asymptomatic primary hyperparathyroidism are few. In 52 patients with asymptomatic primary hy-
perparathyroidism who had iliac crest biopsies and mineral density measured at the radius, spine, and hip, a deficit in cortical but not in cancellous bone was Indeed, the latter tended to be higher than normal, suggesting a protective effect of P T H at skeletal sites with a major cancellous component. The same findings applied whether the mild hyperparathyroidism was or was not associated with stone formation.(54' In a study involving 174 black and white men and women with asymptomatic primary hyperparathyroidism (serum calcium 5 12 mg%), vertebral fractures occurred but the prevalence assessed from radiographs was normal.'ss) There was also no increase in overall fracture prevalence, although bone mass in the forearm was reduced. On the other hand, in 174 women with primary hyperparathyroidism there was an increased prevalence of fracture at the wrist, hip, and spine (Table 4 ) . ( 5 1 1 In these patients with fracture, 72% had a serum calcium I 12 mg% and in 57% this was 5 11.25 mg%. Both cortical bone (metacarpal cortical width) and cancellous bone (Singh grade) were reduced, and the cortical deficit corrected for age was more marked in the elderly patients. The fracture prevalence and bone deficit were highest in those with the lowest absorption of calcium. The rate of cortical bone loss in patients with mild primary hyperparathyroidism is only marginally increased, although parathyroidectomy results in a significant decrease in the rate of loss in the short term.(51)In the long term, however, there is incomplete recovery of the bone defiThese differences in fracture incidence may reflect differences between centers in selection criteria and in methods of measuring fracture occurrence. O n the other hand, they may reflect true geographic variation in fracture prevalence and indicate the need for more studies. TABLE 4. NUMBER OF OBSERVED A N D EXPECTED FRACTURES IN 174 WOMEN WITH PRIMARY HYPERPARATHYROIDISM Skeletal site
Observed
Expected
P
26 23 12 5 4 4
3.9 9.0 1.8 < 0.2 < 0.2 < 0.2
< 0.001 < 0.001 < 0.001 < 0.001
Spine Wrist Femur Wrist and spine Wrist and femur Spine and femur Source: From Ref. 5 1 .
