Exp. Clin. Endocrinol. Vol. 98, No. 3, 1991, pp. 179-184

J. A. Barth, Leipzig

Department of Internal Medicine (Head: Prof. S. Ljunghall), University Hospital, Uppsala/Sweden

L. LIND and S. LJUNGHALL With 2 Figures Summary. In hospitalized patients primary hyperparathyroidism (HPT) and neoplasms account for more than 90% of all hypercalcemias. Measurements of parathyroid hormone, particularly when combined with dynamic tests using calcitonin and EDTA have a high specificity and sensitivity in the differential diagnosis of hypercalcemia but are time-consuming and costly for screening purposes. Most chemical autoanalyzers beside serum calcium also measure serum chloride, phosphate and albumin. In order to evaluate how these simple variables could differentiate between HPT and hypercalcemia due to malignant disorders, 110 measurements from HPT subjects and ill measurements from cancer patients with hypercalcemia were used. Serum chloride was best among the simple variables to separate the two disorders and classified 84% of the hypercalcemic subjects correctly. When serum phosphate and albumin were added giving the formula (serum chloride-84) x (albumin-i 5)/phosphate, only 3% .of the cancer and 4% of the HPT subjects were misclasssified when borderline values (400-500) were excluded (5% of the sample). In conclusion, while other more sensitive and expressive tests exist to establish the cause of hypercalcemia the above mentioned formula is a cheap and easy screening test for a preliminary diagnosis.

Key words: Hyperparathyroidism - Hypercalcemia - Chloride - Phosphate

Introduction

Since the introduction of chemical autoanalyzers it has been discovered that hypercalcemia is a rather common finding being present in about 1% of the adult population (Palmér et al., 1987). Causes of hypercalcemia include primary hyperparathyroidism (HPT), malignancy, sarcoidosis, chronic renal failure, vitamin D intoxication, thyrotoxicosis and some other rare disorders. HPT is characterized by raised levels of parathyroid hormone (PTH) while suppressed levels of this hormone are found in all other causes of hypercalcemia. As PTH promotes a hyperchioremic metabolic acidosis the serum chloride value was earlier suggested to be able to separate cases with HPT from hypercalcemia of other origin (Wills and McGowan, 1964). PTH also has phosphaturic properties and the serum chloride to phosphate ratio was therefore later reported to be more efficient in the separation of hypercalcemic subjects (Palmer et al., 1974; Choo-

Downloaded by: Collections and Technical Services Department. Copyrighted material.

Serum Chloride in the Differential Diagnosis of Hypercalcemia

180

Exp. Clin. Endocrino!. 98 (1991) 3

Kidney cancer Bronchial cancer Multiple myeloma Pancreatic cancer Mammary cancer Primary liver cancer Prostatic cancer Urinary bladder cancer Non Hodgkin lymphoma Cervix cancer Cancer of the tongue Unknown primary tumor site

20% 29% 10% 2% 12% 4% 6% 2% 4% 2% 2% 7%

Kang and Stennet, 1986). The usefulness of the chloride-phosphate ratio has however been questioned (Pak and Townsend, 1976). Later also more sophisticated formulas derived from regression analysis with several other variables have been proposed (Fraiser et al., 1977, Transbøl et al., 1977; Lafferty, 1981). The present study was designed to evaluate to what degree these variables could separate subjects with HPT from hypercalcemia associated with malignancy.

Material and Methods The material for this study was collected at the Gävle Country Hospital. An autoanalyzer which among other variables gives values for serum calcium, albumin, chloride, bicarbonate and phosphate is used in this hospital (SMAC, Technitron Inc., USA). One hundred and ten preoperative serum measurements from 86 patients in whom the diagnosis of HPT could later be confirmed at histopathological examination of removed glands were used together with 111 serum measurements from 53 patients with malignant disorders. Diagnosis of the patients with neoplasms are given in Table 1. All serum measurements included a serum calcium value (corrected for albumin) above 2.60 mmol/l and the measurements were taken at least one week apart. No more than three measurements were allowed from the same individual. To obtain age and sex matched reference values serum measurements from 102 hospitalized normocalcemic patients were used. The controls were matched for age and sex. Statistics: Pearson's correlation coefficients were used for correlation analysis.

Results

The means for the measured variables are given in Table 2. As serum chloride, phosphate and albumin seemed to differ between the hypercalcemic groups cross products and ratios between these variables were calculated and analyzed with regard to the separation of the groups. Of the single variables serum chloride was found to be most effective in separating

the hypercalcemic groups from each other (Fig. 1). If all subjects with serum chloride of 104 mmol/l (5%) were excluded, 92% of the cancer patients and 86% of the HPT patients were correctly classified. The use of the chloride to phosphate ratio did not improve the separation of the groups. Neither did the (chloride-90)/ phosphate ratio, as proposed by Higashi et al. (1985), show any better power in separation.

Downloaded by: Collections and Technical Services Department. Copyrighted material.

Table 1. Causes of hypercalcemia due to malignancy

181

L. LIND and S. LJUNGHALL, Serum Chloride in Hypercalcemia Serum chloride m mo I/I

¿

110

00 8

105

000

00 - - 0000GO 0000000000000000 00008 ... n. 000 00000000 00000000000 000000 00000000000 00000000 0000 00 0000 88 o o

loo

95

90

85

swift

o Cancer

Fig.

..ssss.

.uuu.uuIuuIuIuss uuuIIuuuu-uuuIu

HPT

1. Individual values for serum chloride in HPT and hypercalcemic cancer patients. A

borderline value of 104 mmol/l is indicated.

However, when serum albumin was added in the equation serum chloride x albumin/phosphate and the patients with borderline values of 4O00-44O0 (6%) were excluded, only 6% of the cancer patients and 5% of the HPT patients were misclassified. Furthermore, by subtracting serum chloride and albumin by the lowest values found in the sample, which gives an equation (chloride-84) x (albumin-i 5)/phosphate, and excluding values between 400-500 (5%), only 3% of the cancer and 4% of the HPT patients were misclassified. No significant correlation between serum calcium and chloride was found in the normocalcemic controls, while a positive correlation was seen in the HPT subjects

(r=0.20, p

Serum chloride in the differential diagnosis of hypercalcemia.

In hospitalized patients primary hyperparathyroidism (HPT) and neoplasms account for more than 90% of all hypercalcemias. Measurements of parathyroid ...
52KB Sizes 0 Downloads 0 Views