A New, Vitamin D-Based, Multidimensional Nomogram for the Diagnosis of Primary Hyperparathyroidism

Adrian Harvey, MD; MengJun Hu, MS; Manjula Gupta, PhD; Robert Butler, MS; Jamie Mitchell, MD; Eren Berber, MD; Allan Siperstein, MD; Mira Milas, MD, FACS


Endocr Pract. 2012;18(2):124-131. 

In This Article


Patient Cohort Used in Nomogram Development

In an effort to construct a multidimensional nomogram for the diagnosis of 1°HPT, we based our initial analysis on a cohort of healthy normal subjects. We identified 1,000 consecutive patients with 25-hydroxyvitamin D [25(OH)D] measurements at the Cleveland Clinic in Cleveland, Ohio. Residual blood samples not required for clinical use from these patients are routinely stored for a brief period, in accordance with clinical pathology protocols, and then discarded. We maintained at −20°C the blood samples intended for disposal when they were no longer clinically needed. These samples would be the source of future measurement of total serum calcium and intact PTH concentrations. We considered the incorporation of ionized calcium values in our nomogram development, but this decision would have required recruitment of healthy persons for separate venipuncture and separate specimen processing and testing. Likewise, we considered the incorporation of 1,25-dihydroxyvitamin D levels, but this test is not performed at our institution. We reasoned that the most versatile and relevant nomogram would be one based on variables that are routinely assessed in patients with suspected 1°HPT and available at most institutions.

Using the International Classification of Diseases, Ninth Revision codes to query computerized medical records and confirmation by review of individual patient charts, we excluded from the study those patients with conditions that might affect calcium, 25(OH)D, and PTH metabolism. These exclusion criteria included renal failure, renal insufficiency, metabolic bone disease, any type of cancer, preexisting parathyroid disease, malnutrition, adrenal disorders, transplantation, sarcoidosis, diabetes insipidus, and hyperthyroidism. In addition, patients were excluded if they were taking any of the following medications: lithium, thiazide diuretics, cinacalcet, theophylline, high-dose vitamin D for repletion (ergocalciferol, 50,000 U), calcitriol, and bisphosphonates. We did not exclude or analyze any patients on the basis of race or sex. After this analysis was superimposed on the 1,000 patients whose blood samples were in storage, 222 healthy patients were available for further study. All these patients had normal serum creatinine measurements. Calcium and PTH measurements were performed on sequestered blood samples of these 222 patients. Institutional review board approval was obtained for this portion of the study, as well as for the study of the surgical cohort detailed in a subsequent section.

Biochemical Analysis

All 25(OH)D measurements were performed with a 2-step DiaSorin 25(OH)D assay (DiaSorin Inc., Stillwater, Minnesota). Serum intact PTH was quantified by using the ADVIA Centaur (Siemens USA Healthcare Diagnostics, Deerfield, Illinois) intact PTH assay, a 2-site sandwich immunoassay with antibodies against the N-terminal[1–34] and the 39–84 region of the molecule. Serum calcium measurements were performed with a Roche automated clinical chemistry analyzer (Roche Diagnostics, Mannheim, Germany) and use of a reaction with a cresolphthalein complexone in the presence of 8-hydroxyquinoline. This reaction forms a purple chromophobe, the intensity of which, measured photometrically, is proportional to the calcium concentration.

Mathematical Modeling and Statistical Analysis

In combination with demographic data, the laboratory values from these healthy patients were used to construct a statistical model for the prediction of patient-specific "normal" PTH values. The data used to develop a predictive equation relating PTH to calcium and 25(OH)D were analyzed with use of SAS version 9.2 software (SAS Institute Inc., Cary, North Carolina). The variables to be used in the multivariate study were assessed for colinearity. In addition to the main variables, the data set allowed a test of the possible interaction between 25(OH)D levels and calcium levels as well as the interaction of 25(OH)D levels with the curvilinear relationship between calcium and PTH. None of these terms was significant. The regression model was developed by using stepwise regression methods. Both backward elimination and forward selection with replacement converged to the same model. The final model for defining patient-specific normal PTH values (referred to as the PTH nomogram) consisted of only those variables that were significant (P<.05).

Surgical Validation Cohort

From a prospectively maintained, institutional review board-approved database, we identified a cohort of 351 consecutive patients with 1°HPT who had complete preoperative total serum calcium, PTH, and 25(OH)D levels available. All patients referred to the endocrine surgery clinic have a serum albumin level measured for the purpose of correcting the total calcium level. None of these patients was found to have an abnormal serum albumin value. Measurements of serum albumin were not available for all the "healthy" control subjects, but no diagnoses were noted that would be expected to result in a below-normal level of serum albumin for the healthy cohort. All surgical cohort patients had undergone bilateral cervical exploration with intraoperative PTH guidance and parathyroid excision; pathologic confirmation of diseased glands showed that 226 had a single adenoma, 66 had double adenoma, and 59 had multigland hyperplasia. No patient in the surgical cohort had a "negative" result on neck exploration. The patients underwent assessment by both endocrinologists and surgeons in accordance with currently accepted treatment guidelines for parathyroid disease (summarized in reference 1, from the National Institutes of Health workshops). Moreover, no patient had a preoperative diagnosis of secondary hyperparathyroidism, which would have precluded a decision to proceed with surgical treatment. The 25(OH)D status was comparable among all disease categories found at operation: 25.5 ± 11.4 ng/mL for single adenomas, 26.4 ±11.2 ng/mL for double adenomas, and 26.8 ±13.4 ng/mL for multigland hyperplasia (P = no significant difference). The PTH nomogram was applied to this surgical patient population to determine its performance in the identification of 1°HPT in this setting, where disease was verified by ultimate surgical and histologic standards. In addition, subsets of surgical patients with borderline laboratory values were identified and examined separately to determine how the PTH nomogram performed in these difficult-to-diagnose populations.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.