[Approach to examining hypercalcemia in the clinical laboratory]

The investigation of hypercalcemia is carried out routinely in our laboratory to detect primary hyperparathyroidism (PHPT). In a retrospective study, seven patients with PHPT and fifty-one patients with non-parathyroid hypercalcemia (NPHC) were chosen in a particular year. To obtain a screening index for PHPT, discriminant analysis, using a stepwise variable select method, was applied to eight biochemical parameters in these patients. A discriminant function (F1) was derived from three biochemical parameters and then another discriminant function (F2) was also derived from three biochemical parameters in the F1-positive patients. In combination of these two functions (F1 and F2), the final sensitivity was 100% and specificity was 98% in diagnosing PHPT. This screening method was tested prospectively in fifty-six consecutive specimens of hypercalcemia (PHPT 4, NPHC 52) over the following six months. The result was also satisfactory with a sensitivity of 100% and specificity of 98%. It was proven that our screening method using discriminant functions (F1 and F2) was very useful for diagnosing patients with PHPT from the survey of hypercalcemia. Among these patients with hypercalcemia, the high ratio (54%) of those with malignancy was remarkable. This interesting result required us to investigate potential hypercalcemia, since the serum calcium concentration was masked by a lower level of serum albumin, which was frequently seen in these malignant patients. As the next step, we tried to adjust the serum calcium concentration based on the serum albumin concentration. A formula for adjusting the calcium concentration was derived from a linear structural relationship between calcium and albumin in 6,821 specimens within a +/- 2.5 second principal component score in 7,021 consecutive specimens in whom both calcium and albumin were measured in a particular year; Adjusted Calcium = Calcium - Albumin + 4. After adjustment using this formula, the calcium concentrations were elevated above the upper limit of the reference interval in 320 of 5,203 specimens (6%) within the reference interval and elevated to the reference interval in 1,390 of 1,579 specimens (88%) below a lower limit of reference interval. A prospective study was performed over the following three months. Fifty patients with hypercalcemia were screened using this formula. It was a surprise that thirty-one patients (62%) showed abnormal values after adjustment. These results suggest that calcium adjustment is necessary for interpreting the calcium concentration of patients with a reduced albumin concentration such as patients with malignancy.