JOURNAL OF BONE AND MINERAL RESEARCH Volume 6. Supplement 2. 1991 Mary Ann Liebert, Inc.. Publishers

Differential Diagnosis of Hypercalcemia FREDERIC W. LAFFERTY

ABSTRACT The differential diagnosis of hypercalcemia has expanded to over 25 separate disease states, with primary hyperparathyroidism and malignancy accounting for 8O-9O% of all hypercalcemic patients. Primary hyperparathyroidism comprises the majority of hypercalcemic patients among the ambulatory population, but malignancy accounts for up to 65% of such patients in the hospital. Factors favoring primary hyperparathyroidism include a family history of hyperparathyroidism or multiple endocrine neoplasia, a history of childhood radiation to the head and neck, the postmenopausal state, a history of renal calculi or peptic ulcer, hypertension, the induction of hypercalcemia by thiazides, or an asymptomatic patient with a prolonged, stable mild hypercalcemia. The usefulness of the serum calcium, parathyroid hormone, chloride, phosphorus, serum 25-OHD, and 1,25-(OH),D, and urinary calcium in the differential diagnosis of hypercalcemia is discussed. The pitfalls of an excessive reliance on the serum PTH in diagnosing hyperparathyroidism are stressed. The discriminant values of the serum calcium, chloride, phosphorus, and parathyroid hormone are explored, with the serum parathyroid hormone, chloride, and calcium proving most useful in separating primary hyperparathyroidism from other forms of hypercalcemia. Multivariate discriminant analysis using the serum calcium, phosphorus, and chloride and the hematocrit achieves an accuracy of 95-98% and is the most economical method of identifying hyperparathyroidism. The addition of the amino-terminal or intact PTH assay increases the accuracy to 99% and is essential in the presence of renal insufficiency.

INTRODUCTION of inexpensive, routine serum multiple chemical analysis during the past four decades, the problem of the differential diagnosis of hypercalcemia more often confronts the physician in ambulatory office practice than in the hospital. The purpose of this presentation is to facilitate the differential diagnosis of hypercalcemia in an uncomplicated and economical manner with particular emphasis on identifying the patient with primary hyperparat hyroidism. Hypercalcemia results when the sum of the calcium coming from bone resorption (BR), the net intestinal absorption, and from soft tissue deposits exceeds the quantity of calcium being deposited in bone (BF) and soft tissue and excreted in urine."l

W

ITH THE A D V E N T

BR + net intestinal Ca absorption + soft tissue Ca release > BF + soft tissue deposition + urinary Ca tserum Ca

-

With good renal function and suppressed parathyroid hormone secretion, one may excrete in excess of lo00 nig calcium per day in the urine, but with a creatininc clearance of less than 40 ml/minute the excretion may be limited to less than 200 rng/day.'l)

DIFFERENTIAL DIAGNOSIS AND CLINICAL PRESENTATION The diagnostic spectrum of hypercalcemia has steadily expanded over the past 30 years to a current list in excess of 25 different entities, as shown in Table I . The mechanism of hypercalcemia in these disease states has become more clearly understood with the development of assays for parathyroid hormone, parathyroid-related protein, 25hydroxyvitamin D (25-OHD), and 1,25-dihydroxyvitamin D [ 1,25-(OH),D] but is still incompletely understood for many malignancies, myxedema, familial hypocalciuric hypercalcemia, thiazides, lithium, milk alkali syndrome, Wil-

Department of Medicine, University Hospitals of Cleveland, and Case Western Reserve School of Mcdicinc. C'lcvcland, Ohio.

S5 1

LAFFERTY TABLE 1. HYPERCALCEMIA: DIFFERENTIAL DIAGNOSIS Endocrine Hyperparathyroidism Primary Secondary (postrenal shutdown or transplant, postpancreatitis) Tertiary (uremia, malabsorption) Familial hypocalciuric hypocalcemia (FHH) Hyper- and hypothyroidism Adrenal insufficiency Pheochromocytoma Vipoma Malignancy Humoral hypercalcemia of malignancy [PTHrP, 1,25(OH),D, PTHI Metastatic bone disease (local osteolytic factors) Exogenous agents Thiazides Vitamin D Vitamin A Milk alkali Lithium Aluminum Beryllium Theophylline Granulomatous disease Sarcoidosis Tuberculosis Leprosy Fungal Immobilization Miscellaneous Post-acute renal failure with rhabdomyolysis Infantile hypercalcemia (William's syndrome) Advanced chronic liver disease Disseminated cytomegalovirus in AIDS

ham's syndrome, rhabdomyolysis, and advanced liver disease. The importance of the local production in bone of prostaglandins, interleukins, tumor necrosis factors, and transforming growth factor remains to be determined.I3) Primary hyperparathyroidism and malignancy account for 8O-9OVo of hypercalcemia states seen today. In the ambulatory setting hyperparathyroidism accounts for 50-60% of h y p e r c a l ~ e m i a , ~in' ~ the ~ ~ hospital for 27% or less.'6.') By contrast, malignancy accounts for 31% or less of ambulatory hypercalcemia but for up to 65% of that seen in the hospital (Table 2). Carcinomas of the bronchus, breast, head and neck,, urogenital tracts, and multiple myeloma account for 75% of the hypercalcemia in malignancy.") A humoral mechanism with an elevated nephrogenous cyclic AMP and serum parathyroid-related protein (PTHrP) or an elevated serum 1,25-(OH),D may be found in up to 80% of malignancies with hypercalcemia.'") Local osteolytic factors predominate in the remaining 20070, including multiple myeloma and carcinoma of the breast.I3) One must also be cognizant of the concurrence of primary hyperparathyroidism and malignancy, especially with carcinomas of the breast, thyroid, gastrointestinal tract, and genitourinary tract. 1 A careful history is the basis for evaluating all patients with hypercalcemia even though a cause seems readily apparent. The use of thiazides, large doses of vitamins A or D, lithium, or large amounts of calcium carbonate or theophylline must be eliminated before proceding with more complex tests. Immobilization, recent renal failure associated with muscle injury, and a history of tuberculosis or fungal infection may be quickly determined. A family history of hypercalcemia suggestive of familial hypocalciuric hypercalcernia or multiple endocrine neoplasia are important clues. Finally, a long-term history of stable, mild hypercalcemia without symptoms or weight loss is most consistent with a diagnosis of primary hyperparathyroidism,

TABLE2. FREQUENCY OF HYPERCALCEMIC STATESIN AMBULATORY AND HOSPITAL POPULATIONS~ Ambulatory Cause

N

=

Ref. 4: 61/26,000 (%)b

Hospital Ref. 5:

N

= 95/15,903

Hyperparathyroidism Malignancy Hypervitaminosis D Milk alkali syndrome Hyperthyroidism Hypothyroidism Addison's disease Sarcoidosis Thiazides Renal transplant or dialysis Unknown ~~

=Frequency is indicated by N = number per population screened. bPrevalence of hypercalcemia was 0.23% in the ambulatory population. CPrevalence of hypercalcemia was 0.60% in the ambulatory population.

Ref. 6: (%)C

N

=

57(%)

N

Ref. 7: 469 (%)

=

27 54 2.3 1 -

0.2 -

14.5 1

Diagnosis Primary hyperparat hyroidism Familial hypocalciuric hypercalcemial

TABLE3. USEFUL TESTSIN ~

N( 1)

Nt

1N

I

I

Serum P

Intact serum PTH

Serum CI

N( 1)

THE ~~~

Urinary CU

DIFFERENTIAL DIAGNOSIS OF HYPERCALCEMIA~ ~~

Serum 2.5-OHD,

tN

N

1N

N

~

Serum I,2S-(OH),D3

~

f

t

N(1)

N

It

1(N)

Thyrotoxicosis' I J 1

1

Nt

1

N

It

1N

1

1(1)

N

1

1

Nt

1

1

It

tf

Nf

1

(N)

N

t

I

1

1

Milk alkali syndr~me"~~ Lithium'20'

1

Vitamin D intoxication'18)

N1

N I

Ectopic production of 1,25(OH),D similar to granulomatous diseases Elevated NcAMP suggests humoral mechanism Hypercalcemia resolves within 2 weeks of stopping thiazides Excess 1,25-(OH),D ingestion causes 1 25-OHD and t 1,25(OHLD Elevated serum creatinine and bicarbonate Seen in 10% of patients on long-term lithium Dialysis, TPN, biliary tract disease Unmasked or aggravated by small doses of vitamin D

1

1

1

N

tt

tt

1

Nt1

1

f

NJ

Nf

Nl

N

1

N N N

1

It

IN

I

Nt

N1

I

f

Nt

1

1

1

I

1

N

IN

NI

N

1

tt

Nt

1

Aluminum intoxication(1" Sarcoidosis and 1 granulomatous diseases(z2' Imm~bilization'~~' 1

1

~~~~~~

Comments

NcAMP

t

Radiographic evidence of osteitis fibrosa cystica seen in only 5% Ratio of renal calcium clearance to creatinine clearance less than 0.01 Hypercalcemia may mask symptoms of thyrotoxicosis Confirmed by a high serum PTHrP and a low-normal or low PTH

I I. I 1I

f

N

tt

Nt

1

1

Humoral hypercalcemia of malignancy (PTHrP)"'] Lymphoma and Hodgkins"" Metastatic bone disease' 7 , 1 6 ) Thiazides'I7]

Nt

1

I

I

Idiopathic hypercalcemia of infancy' 2 4 ) Advanced chronic liver disease[15' Postacute renal failure with rhabdomyolysisl161 at,

Develops within 1-3 months in high bone turnover states Elfin facies and growth retardation, hypersensitive to vitamin D Seen in 2% of advanced cases considered for transplantation

N

I

1 Muscle uptake of m T c medro-

nate on bone scan

high; I , low; N. normal; parentheses indicate occasionally. The first symbol represents the more common finding.

LAFFERTY

s54

familial hypocalciuric hypercalcemia, and occasionally sarcoidosis. A short-term picture of a rising serum calcium, weight loss, or fever point toward malignancy, tuberculosis, fungal infection, cytomegalovirus in acquired immune deficiency disease, and adrenal insufficiency. The physical examination is usually normal in primary hyperparathyroidism and most other forms of hypercalcemia. Notable exceptions are the presence of tumor masses or nodes in malignancy, hepatosplenomegaly in malignancy, granulomatous disease or chronic vitamin A intoxication, hyperpigmentation in Addison's disease, and a goiter in thyrotoxicosis. Hypertension increases the likelihood of primary hyperparathyroidism, but a pheochromocytoma must be ruled out.'1o)Of course, severe hypercalcemia may lead to dehydration, cachexia, weakness, and metastatic hypercalcemia in all forms of hypercalcemia. Factors favoring the diagnosis of primary hyperparathyroidism include a positive family history of hyperparathyroidism or multiple endocrine neoplasia, a history of childhood radiation of the head and neck, a postmenopausal woman, a history of renal calculi or peptic ulcer disease, hypertension, the induction of hypercalcemia by thiazides, or an asymptomatic patient with a prolonged, stable mild hypercalcemia and a normal physical examination.

TABLE 4. PRIMARY HYPERPARATHYROIDISM: PITFALLS LABORATORY A low-Ca and high-P diet may obscure mild hypercalcemia The serum intact PTH may be high-normal in 10-15% primary hyperparathyroidism A high Ca and PTH seen in 20% of FHH A high Ca and P T H seen with lithium hypercalcemia A high intact P T H is rarely seen in malignancy (true ectopic PTH syndrome) A high Ca and CI may be seen in thyrotoxicosis Renal failure may destroy value of serum chloride, phosphorus, 1,25-(OH),D, urinary calcium, hematocrit, NcAMP, and multivariate analysis but not the value of serum intact or N-terminal PTH

X-RAY AND LABORATORY TESTING All patients with hypercalcemia should have a complete blood count, routine serum chemical profile, and chest x-ray. One can also make a case for routinely determining the serum free thyroxine index and high-sensitivity thyroidstimulating hormone (TSH) since the signs and symptoms of thyrotoxicosis may be obscured by hypercalcemia, especially among older patients. Routine skeletal x-ray surveys and nuclear bone scans are unnecessary unless there is a suspicion of malignancy or the serum alkaline phosphatase is elevated in the presence of a normal serum gamma-glutamyltransferase. Useful laboratory tests in the differential diagnosis of hypercalcemia have expanded from the serum calcium, phosphorus, and alkaline phosphatase and the 24 h urinary calcium in 1948 to currently include the serum ionized calcium, immunoradiometric intact PTH, chloride, 25-OHD, 1,25-(OH),D, osteocalcin, and the urinary cyclic AMP. Table 3 is a composite of the prevailing results of many of these tests from the literature as well as the author's personal experience. The blanks represent a lack of sufficient data. Additional laboratory tests are needed according to the patient's presentation. The hypercalcemic patient with hypertension should have a measurement of the urinary catecholamines to rule out a pheochromocytoma, and the patient with intractable diarrhea and a dilated gallbladder needs measurement of the serum vasoactive intestinal polypeptide (VIP) to determine if a vipoma of the pancreas is present. The presence of acute renal failure following massive muscle injury requires measurement of the serum creatine kinase, and the patient with weakness, hypotension, or a low serum sodium needs determination of the morning serum cortisol. Following the completion of the history just outlined, physical examination, and basic radiologic and laboratory tests, the clinician may still be faced with a differential diagnosis of primary hyperparathyroidism, familial hypocalcemic hypercalcuria, occult granulomatous disease, and occult malignancy. The task is made easier by an elevated serum intact PTH, but elevated levels may be seen in 20% of patients with familial hypocalcemic hypocalcuria and most lithium-induced hypercalcemia.(11,12.zo' A 24 h uri-

TABLE 5 . RADIOIMMUNOASSAYS OF P T H R P

P T H r P fragment Detectable in normals, N (To) Elevated P T H r P All malignant hypercalcemias, N (To) H H M syndrome, N (To) Primary hyperparathyroidism, N ('7'0) aSolid tumors exclusive of breast and prostate. bSolid tumors without skeletal metastases. CDefined by an elevated urinary cyclic AMP.

Ref. 30

Ref. 31

Ref. 32

Ref. 8

1-37 5/17 (32)

1-34 2/39 ( 5 )

1-34 43/48 (90)

1-74 28/60 (47)

36/65 ( 5 5 ) 30/42 (71)a 1/16 (6)

19/38 (50)

17/36 (47) 10/19 (53)b 0/8 (0)

30/38 (79) 25/30 (83)c 0/13 (0)

4/20 (20)

DIFFERENTIAL DIAGNOSIS OF HYPERCALCEMIA

SERUM CALCIUM MG%

r20 I

:

I b18 I I

I

I-1 6

.. ..*.. ........... ............. ........... .................... ...... I

0.

A

0 AX A

A

8

.X

XAXXX

A A.

XAAXX

B~.K~~x ox xxxx

X XXAOA

AA

OX AX .X .XXX

0

I I I I

0.

AA

.

SERUM C H L O R I D E MEQ/L

-1 1 0 A 0

I

A

I

0

105

0..

0

0..

A

xxxx

I

0 0 . 0 0 . .

-1

xxx

03d

A.

000.00

0 0.0

I

0

XAOXX 0

X

A.AX

-

AX.

1I

00

110

XAXX

-100-

X.

- x

AAXAX

I

X0A.X

-95-

AAXAX X

I

I I

X.A.

X

X.AX

I

-90I I

1I 6 PRIMARY OTHER HYPERPARATHYROlD HYPERCALCEMIAS

xx A

I

I

I 0

X X

I

PRIMARY HYPERPARATHY ROlD

OTHER HYPERCALCEMIAS ~~

~

shown by dashes. The best discriminant level is 12.5 mg/dl (3.12 mmol/liter). (X) Metastatic bone disease; (A)humoral hypercalcemia of malignancy; (0)thyrotoxicosis. (From Lafferty FW 1988 Disorders of calcium metabolism and parathyroid function. In: Mendelsohn G (ed.) Diagnosis and Pathology of Endocrine Diseases. Copyright 1988, J.B. Lippincott Co.)

FIG. 2. Serum chloride levels in hyperparathyroidism versus other forms of hypercalcemia. The mean of each group is shown by the dashes. The best discriminant level was 103 mEq/liter (103 mmol/liter): (0)diuretics or vomiting; (X) metastatic bone disease; (A)humoral hypercalcemia of malignancy; (0)thyrotoxicosis. (From Lafferty FW 1981 Primary hyperparathyroidism. Arch Intern Med 141:1761. Copyright 1981, American Medical Association.)

nary calcium of less than 100 mg, or more specifically a ratio of the renal clearance of calcium to the clearance of creatinine of less the 0.01, indicates familial hypocalciuria hypercalcemia. Finally, in the author’s experience 10-15% of patients with primary hyperparathyroidism may have an intact serum PTH in the upper normal range so that the lack of a high value may not rule out the diagnosis. The importance of the intake of calcium and phosphorus to the serum calcium and parathyroid hormone levels deserves emphasis. A low daily phosphorus intake of

430 mg may increase the serum calcium by 0.3-0.7 mg/dl in hyperparathyroidism, whereas no such rise occurs among subjects.(z71Conversely, a high daily phosphorus intake of 3000 mg may reduce the serum calcium in hyperparathyroidism and other forms of hypercalcemia. Recently it has been reported that an oral calcium load of I500 mg causes a mean rise in the serum calcium of 0.5 mg/dl within 2 h in hyperparathyroidism, with only a 0.3 mg/dl rise in normal subjects. The serum amino-terminal P T H fell by 73% in normal subjects, whereas i t fell by

FIG. 1. Serum calcium levels in hyperparathyroidism versus other hypercalcemic states. The mean of each group is

LAFFERTY

S56

TABLE6. COSTEFFECTIVENESS OF LABORATORY TESTING HYPERPARATHYROIDISM VERSUS OTHERHYPERCALCEMIC STATES Tests Single Serum calcium Intact PTHb Serum chloride Serum phosphorus Hematocrit Multiple Ca, CI, P, Hct Ca, CI, P , Hct, PTH

Discriminant levels 12.5 mg/dl or 3.12 mmol/liter >65 pg/ml 2 103 mEq/liter 5 2 . 9 mg/dl or I0.94 mmol/liter 2 39% 5 5

Retrospective accuracy (To)

Cost to patient (CI.S.) dollarsa

84

8

85-90 91 72

68 8 8

72

4

98 99

30 98

C o s t s represent 1990 charges in Cleveland, Ohio. bBased on experience with Nichols Laboratory two-site immunoradiometric assay of intact parathyroid hormone. Essential test in renal insufficiency (serum creatinine > 2.5 mg/dl).

only 21 070 in hyperparathyroidism.'z81Table 4 outlines some of the laboratory pitfalls in the diagnosis of primary hyperparathyroidism. When the serum immunoradiometric intact PTH is low normal or subnormal in the hypercalcemia patient, one is often faced with a differential diagnosis between occult granulomatous disease and malignancy. A high serum 1,25-(OH),D, suggests granulomatous disease, but patients with lymphoma and Hodgkin's disease may also have elevated levels. Both conditions may present with hepatosplenomegaly and adenopathy on abdominal computed tomography (CT), and both may respond to suppressive doses of prednisone. Without the typical changes of granulomatous disease on chest x-ray, only a biopsy can distinguish between these diseases. The laboratory distinction between primary hyperparathyroidism and the humoral hypercalcemia of malignancy (HHM) has been facilitated by the serum immunoradiometric intact PTH in which patients with hypercalcemia of malignancy consistently have low or low normal levels and patients with primary hyperparathyroidism have high or high normal levels of PTH."') The presence of a high serum chloride and 1,25-(OH),D also favors hyperparathyroidism over malignancy, but these tests are not always reliable discriminators. The new assay of the serum parathyroid hormone related protein (PTHrP) facilitates the differentiation of the humoral hypercalcemia of malignancy from not only primary hyperparathyroidism but, more importantly, from other forms of hypercalcemia in which the serum PTH is suppressed. T o date four assays for serum PTHrP have been d e ~ c r i b e d , ( ~ ,but ' ~ -problems ~~~ still exist (Table 5). In two of the a ~ s a y s ' ~ OP.T~H~ r)P has been de-

tected in primary hyperparathyroidism (6-20%). Only the two-site assay of the 1-74 amino-terminal fragment described by Burtis et a1.(8)found elevated PTHrP levels in a high percentage (83%) of cases of HHM and no elevated level among 13 patients with primary hyperparathyroidism. A commercially available PTHrP two-site immunoradiometric assay using the Burtis antibodies is now available, but its sensitivity and specificity in widespread clinical use have yet to be determined. The rare true ectopic P T H syndrome in which a malignant tumor secretes PTH rather than PTHrP has long been suspected but only recently documented. This was strongly suggested by the finding of elevated serum PTH levels plus mRNA coding for PTH in small cell lung cancers from two patients with h y p e r c a l ~ e m i a . ( ~Nuss~.~~) baum et al.'.35)recently reported secretion of PTH by an ovarian carcinoma in which the serum P T H rapidly fell following resection of the tumor.

MULTIVARIATE DISCRIMINANT ANALYSIS The most cost efficient use of laboratory tests in separating primary hyperparathyroidism from other forms of hypercalcemia can be addressed by the use of multivariate discriminant analysis using the serum calcium, chloride, and phosphorus and the hematocrit. This was originally described by the author in 1981(361by comparing the laboratory values among I 0 0 consecutive cases of surgically proven primary hyperparathyroidism with 64 cases of

DIFFERENTIAL DIAGNOSIS OF HYPERCALCEMIA

D I SC R I MI N A N T

S C ORE

I I I

I

10 I

I I

4 --

1

0

0I I I

0

A

I I I

-5

-----A X

0

-10 I I I

-15 I I I

-20 I

.X XXXA

0.22Hct

+

0.76CI - 1.5Ca,,

-

1.9P

-

77.4

0 . XA.

xxxxx .XX XX.

A h A

:i

i:

AA X

xx X X

A

. X

1

PRIMARY HYPERPARATHY ROIDS

of accuracy for each test are shown in Table 6. In the absence of vomiting and diuretic use, the serum chloride proved to be the most reliable discriminator, with only 9% misclassifications. Thus, some have referred to the serum chloride as “the poor man’s PTH assay.” Thyrotoxicosis and 20% of HHM cases may also show serum chloride levels of 103 mEq/liter or higher. Serum calcium levels above 13.5 mg/dl are unusual in hyperparathyroidism but are almost the rule in untreated malignancies with hypercalcemia. A discriminant function using the serum chloride, calcium, and phosphorus and the hematocrit was derived so that a positive score indicates primary hyperparathyroidism and a negative score another type of hypercalcemia with an overall accuracy of 95%, and an accuracy of 98% in the absence of vomiting, renal failure, or the use of diuretics.

ox

I

I I

S51

OTHER HYPERCACCEMIAS

FIG. 3. Discriminant scores using the serum calcium, phosphorus, chloride, and hematocrit in hyperparathyroidism versus other hypercalcemias. (0) Diuretics or vomiting; (X) metastatic bone disease; (A)humoral hypercalcemia of malignancy; (0)thyrotoxicosis. (From Lafferty FW 1981 Primary hyperparathyroidism. Arch Intern Med 141:1761. Copyright 1981, American Medical Association.)

other forms of hypercalcemia including 31 cases of metastatic bone disease, 20 cases of HHM or pseudohyperparathyroidism, 4 cases of vitamin D intoxication, 3 cases of thyrotoxicosis, 2 cases of thiazide-induced hypercalcemia, and I case each of milk alkali syndrome, sarcoidosis, Addison’s disease, and marked hyperglycemia. The best discriminant levels of serum calcium (Fig. l), phosphorus, and chloride (Fig. 2), the hematocrit, and the percentages

Calcium excess Caex is the difference between the total serum calcium expressed in milligrams per deciliter in a given patient and the upper limit of normal for a given laboratory. The hematocrit is expressed as a percentage, the chloride in millequivalents per liter, and the phosphorus in milligrams per deciliter. The discriminant scores of 100 cases of primary hyperparathyroidism and 64 cases of other hypercakemias are contrasted in Fig. 3. All 5 patients with primary hyperparathyroidism who were misclassified with a negative score were taking diuretics. The validity of this discriminant function using the serum calcium, chloride, and phosphorus and the hematocrit was tested prospectively by Keller and Keller”’) among 37 patients with verified primary hyperparathyroidism and found to be correct in 35 for a 95% overall accuracy. This formula is not applicable to separating primary hyperparathyroidism from either normal subjects or patients with secondary hyperparathyroidism. The addition of the serum PTH increases the accuracy of multivariate analysis to 99%. Since renal insufficiency alters the serum phosphorus and chloride and lowers the hematocrit, multivariate analysis using these factors is unreliable with serum creatinine values above 2.5 mg/dl. One must rely on the combination of an elevated amino-terminal or immunoradiometric intact serum PTH and hypercalcemia to make the diagnosis of primary (or tertiary) hyperparathyroidism in renal failure. The use of multivariate discrimination analysis of laboratory tests in the differential diagnoses of hypercalcemia was first described by Fraser et a1.(”8’in 1971 and has subsequently been explored by seven other groups (Table 7).(39-45) The final proof of the validity of these multivariate discriminant formulas rests with their prospective accuracy, which has been reported in half of these studies. Only Benson reported discriminant analysis between primary hyperparathyroidism and normal subjects with a 92% accuracy using the albumin-corrected serum total calcium, the serum ionized calcium, and an assay to the midportion (44-68) of human PTH.‘45’

LAFFERTY

S58

TABLE 7. MULTIVARIATE DISCRIMINANT ANALYSES Hyperparathyroid N

Reference

68 I58 78 36

38 39 40 41 36

100

100 76 40 16 106 88 91

42 43 44 45 ~

Other hypercalcemia

Retrospective accuracy

N

Discriminant factorsa

19 32 23 48 64

P, BUN, AP, C1, HCO; P, BUN, AP, CI, HCO;, ESR Ca, P, Ccr, UCa P, TRP, CI, HCO;, AP, Ca, Cr CI, Ca, P, Hct C1, Ca, P, Hctb PTH, CI, Ca, P, Hctb Alb, Ca, AST, CI, HCO; PTH, AP, CI, Ca Alb, PTH, C1 Cam, Ca*+,PTH Cam, Ca2+,PTH

64 20 64c 14 74c 46 88 Normals

(YO)

Prospective accuracy N (%)

90

123/140 (88)(”’

93 94 96 95 98 99

-

23/26 (88) 34/34 (100) 35/37 (95)(37’

90 97 94 81 92

~

aListed in order of decreasing discriminatory value. Symbols include AP, serum alkaline phosphatase; ESR, erythrocyte sedimentation rate, C,,, creatinine clearance; U,,, 24 h urinary calcium; AST, serum aspartate transaminase; Alb, serum albumin; Ca,,, total serum Ca corrected for albumin; and Ca”, serum ionized calcium. bPatients with vomiting or diuretics excluded. CMalignancy cases only.

REFERENCES I . Lafferty FW 1988 Disorders of calcium metabolism and parathyroid function: Lboratory evaluation. In: Mendelsohn G (ed.) Diagnosis and Pathology of Endocrine Disease. J.B. Lippincott, Philadelphia, pp. 119-137. 2. Lemann J Jr, Adams ND. Gray RW 1979 Urinary calcium excretion in human beings. N Engl J Med 301535-541. 3. Mundy GR 1989 Calcium Homeostasis: Hypercalcemia and Hypocalcemia. Martin Dunitz, London, pp. 37-181. 4. Boonstra CE, Jackson CE 1965 Hyperparathyroidism detected by routine serum calcium analysis: Prevalence in a clinic population. Ann lntern Med 63:468-474. 5. Christensson T , Hellstrom K, Wengle B, Alveryd A, Wikland B 1976 Prevalence of hypercalcemia in a health screening in Stockholm. Acta Med Scand 200:131-137. Lafferty FW 1966 Pseudohyperparathyroidism. Medicine (Baltimore) 45:247-260. Fisken RA, Heath DA, Bold AM 1980 Hypercalcemia-a hospital survey. Q J Med 49:405-418. Burtis WJ, Brady TG, Orloff J J , Ersbak JB, Warrell R P Jr, Olson BR, Wu TL, Mitnick ME, Broadus AE, Stewart AF 1990 lmmunochemical characterization of circulating parathyroid hormone-related protein in patients with humoral hypercalcemia of cancer. N Engl J Med 322:1106-1112. 9. Kaplan L, Katz AD, Ben-lsaac C , Massry SG 1971 Malignant neoplasms and parathyroid adenoma. Cancer 28:401-407. 10. Rosenthal FD, Roy S 1972 Hypertension and hyperparathyroidism. Br Med J 4:396-397. 1 1 . Marx SJ, Attie MF, Levine MA, Spiegel AM, Downs RW Jr, Lasker RD 1981 The hypocalcuric or benign variant of familial hypercalcemia: Clinical and biochemical features in fifteen kindreds. Medicine (Baltimore) 60:397-412. 12. Firck AF, Kao PC, Heath 111 H 1991 Plasma intact parathyroid hormone (PTH) and PTH-related peptide in familial be-

nign hypercalcemia. J Clin Endocrinol Metab 72541 -546. 13. Burman KD Monchik JM, Earl1 JM, Wartofsky M D 1976 Ionized and total serum calcium and parathyroid hormone in hyperthyroidism. Ann Intern Med 84568-671. 14. Stewart AF, Horst R, Deftos LJ, Cadman EC, Lang R, Broadus AE 1980 Biochemical evaluation of patients with cancer-associated hypercalcemia: Evidence for humoral and nonhumoral groups. N Engl J Med 303:1377-1383. 15. Breslau NA, McGuire JL, Zerwekh JE, Frenkel EP, Pak CYC 1984 Hypercalcemia associated with increased serum calcitriol levels in three patients with lymphoma. Ann Intern Med 1OO:l-7. 16. Mundy GR. lbbotson KJ, D’Souza SM, Simpson EL, Jacobs JW, Martin TJ 1984 The hypercalcemia of cancer: Clinical implications and pathogenic mechanisms. N Engl J Med 310: 1718-1727. 17. Van den Berg C J , Tucker RM, Dousa T P 1982 Idiopathic hypercalcuria: Hydrochlorothiazide decreases urinary calcium without altered renal response to parathyroid hormone. J Clin Endocrinol Metab 55:23-26. 18. Holmes RP, Kummerow FA 1983 The relationship of adequate and excessive intake of vitamin D to health and disease. J Am Coll Nutr 2:173-199. 19. Orwoll ES 1982 The milk-alkali syndrome: Current concepts. Ann Intern Med 97:242-248. 20. Stancer HC, Forbath N 1989 Hyperparathyroidism, hypothyroidism, and impaired renal function after 10 to 20 years of lithium treatment. Arch Intern Med 149:1042-1045. 21. Boyce BF, Elder HY, Elliot HL, Fogelman I , Fell GS, Junor BJ Beastall G , Boyle IT 1982 Hypercalcemic osteomalacia due to aluminum toxicity. Lancet 2:1009-1013. 22. Koide Y , Kugar N, Kimura S, Fujita T, Yamashita N, Hiramoto T , Sukegawa J, Ogata E, Yamashita K 1981 lncreased 1,25-dihydroxy-cholecalciferolas a cause of abnormal calcium metabolism in sarcoidosis. J Clin Endocrinol Metab 52: 494-498. 23. Stewart AF, Adler M, Byers CM, Segre GV, Broadus AE

DIFFERENTIAL DIAGNOSIS OF HYPERCALCEMIA

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

1982 Calcium homeostasis in immobilization: An example of resorptive hypercalcuria. N Engl J Med 306:1136-1140. Taylor AB, Stern P H , Bell NH 1982 Abnormal regulation of circulating 25-hydroxyvitamin D in the Williams syndrome. N Engl J Med 306:972-975. Gerhardt A, Greenberg A, Reilly J J Jr, Van Thiel DH 1987 Hypercalcemia: A complication of advanced chronic liver disease. Arch Intern Med 147:274-277. Akmal M, Bishop JE. Telfer N, Normal AW, Massry SG 1986 Hypocalcemia and hypercalcemia in patients with rhabdomyolysis with and without acute renal failure. J Clin Endocrinol Metab 63:137-142. Eisenberg E 1968 Effects of varying phosphate intake in primary hyperparathyroidism. J Clin Endocrinol Metab 28: 65 1-660. Tohme JF, Bilezikian J P , Clemens TL, Silverberg SJ, Shane E, Lindsay R 1990 Suppression of parathyroid hormone secretion with oral calcium in normal subjects and patients with primary hyperparathyroidism. J Clin Endocrinol Metab 70:951-956. Blind E, Schmidt-Gayk H, Scharla S, Flentje D, Fischer S, Gohring U , Hitzler W 1988 Two-site assay of intact parathyroid hormone in the investigation of primary hyperparathyroidism and other disorders of calcium metabolism compared with a midregion assay. J Clin Endocrinol Metab 67: 353-360. Budayr AA. Nissenson RA, Klein RF, Pun KK, Clark O H , Dinh Diep BS, Arnaud C , Strewler GJ 1989 Increased serum levels of a parathyroid hormone-like protein in malignancyassociated hypercalcemia. Ann Intern Med 1112307-812. Henderson JE, Sherstek C , Kremer R, Rabbani SA, Hendy GN, Glotzman D 1990 Circulating concentrations of a parathyroid hormone-like peptide in malignancy and in hyperparathyroidism. J Bone Miner Res 5:105-113. Kao PC, Klee GG, Taylor RL, Heath H 111 1990 Parathyroid hormone-related peptide in plasma of patients with hypercalcemia and malignant lesions. Mayo Clin Proc 651399-1407. Schmelzer HJ, Hesch RD, Mayer H 1985 Parathyroid hormone and PTH mRNA in a human small cell lung cancer. Cancer Res 99:88-93. Yoshimoto K, Yamasaki R, Sakai H. Tezuka U, Takahashi M, Iizuka M, Sekiya T. Saito S 1989 Ectopic production of parathyroid hormone by small cell lung cancer in a patient with hypercalcemia. J Clin Endocrinol Metab 68:976-981. Nussbaum SR, Gaz RD, Arnold A 1990 Hypercalcemia and

S5Y

36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

ectopic secretion of parathyroid hormone by an ovarian carcinoma with rearrangement of the gene for parathyroid hormone. N Engl J Med 323:1324-1328. Lafferty FW 1981 Primary hyperparathyroidism: Changing clinical spectrum, prevalence of hypertension. and discriminant analysis of laboratory tests. Arch Intern Med 141: 1761- 1766. Keller B, Keller H 1986 A distribution-free, multivariate discriminating method: Initial experience with discrimination 0 1 patient groups by 2 or more clinical chemical parameters. .I Clin Chem Clin Biochem 24:60-609. Fraser P , Healy M, Rose N, Watson L I971 Discriminant functions in differential diagnosis of hypercalcemia. Lancet 1:1314- 1319. Fraser P, Healy M, Watson L 1976 Further experience with discriminant functions in the differential diagnosis of hypercalcemia. Postgrad Med J 52:254-257. Transbdl I , Schonau Jorgensen F, Hornum I , Keiding N 1977 Hypercalcemia discrimination index: A multivariatc analysis of parathyroid function in 107 hypercalcemic patients. Acta Endocrinol (Copenh) 86:768-783. LoCascio V, Vallaperta P , Adami S, Coniinacini L , Galvanini G , Bianchi l, Ferrari M, Scuro LA 1978 Discriminant analysis in the differential diagnosis of hypercalcemia. Clin Endocrinol ( 0 x 0 8:349-356. Fisken RA, Heath DA, Somers S 1981 Hypercalcemia in hospital patients: Clinical and diagnostic aspects. Lancet 1:202207. Wong ET, Freier EF 1982 The differential diagnosis of hypercalcemia: An algorithm for more effective use of laboratory tests. JAMA 247:75-80. Boyd JC, Ladenson J H 1984 Value of laboratory tests in the differential diagnosis of hypercalcemia. Am J Med 77:863887. Benson L, Ljunghall S, Groth F, Falk H , Hvarfner A, Rastad J, Wide L, Akerstrom G 1987 Optimal discrimination of mild hyperparathyroidism with total serum calcium, ionized calcium and parathyroid hormone measurements. Ups J Med Sci 92:147-176.

Address reprint requests to: Frederic W. L.ufferty, M.D. 1611 South Green Road Cleveland, OH 44121

Differential diagnosis of hypercalcemia.

The differential diagnosis of hypercalcemia has expanded to over 25 separate disease states, with primary hyperparathyroidism and malignancy accountin...
643KB Sizes 0 Downloads 0 Views