In 1963, Berson and Yalow developed the first radioimmunoassay for PTH [24] (Fig. 42.1). Their early first-generation (competitive) assay employed polyclonal sera from guinea pigs and rabbits raised against bovine PTH [24]. Because of the assay’s low sensitivity, circulating PTH levels had to be significantly elevated (e.g., patients with hyperparathyroidism) to be detected. A decade later, Arnaud and colleagues (Mayo Clinic) enhanced the assay’s sensitivity for human PTH by using antibodies raised against porcine (rather than bovine) PTH, which allowed for measurement of PTH levels within the normal human range [25].

However, because of high cross-reactivity with small C-terminal fragments and insufficient sensitivity for measuring intact hormone, first-generation assays were replaced by two-site (noncompetitive) immunometric or “sandwich assays” [26]. Second-generation or “intact” PTH assays employ a C-terminal capture antibody (e.g., epitopes 39–44) bound to a solid phase and a second enzyme-labeled N-terminal reporter or signal antibody (e.g., epitopes 13–34) [27] (Fig. 42.2).

While initially thought to measure only intact hormone, later characterization of the assay demonstrated that the detection antibodies cross-react with large C-terminal fragments, which have truncated N-terminal structures (e.g., PTH_7–84) [12]. Depending on the assay, large C-terminal fragments can account for up to 10–30 % of immunoreactive PTH in patients with normal renal function, and up to half in patients with renal failure [6, 13]. Despite these limitations, intact assays are still widely used in many clinical laboratories, and reliable reference ranges have now been established in many patient populations.

To eliminate cross-reactivity with all C-terminal fragments, third-generation or “bioactive” (“whole” or “bio-intact”) PTH assays were developed. Bioactive PTH assays employ a similar capture antibody compared to intact assays but use detection antibodies directed against epitopes located at the extreme N-terminal region (e.g., epitopes 1–4) [28–31] (Fig. 42.2).

The detection antibodies were later demonstrated to cross-react with yet another PTH molecular form, which had not previously been appreciated [32]. This newly discovered molecular form, referred to as N-PTH, is not a PTH fragment but rather a posttranslational modification (possibly a phosphorylated serine at position 17) of PTH_1–84 [28, 30, 31, 33]. Because the modification exists within epitopes 15–20, intact assays do not detect N-PTH. While the physiological relevance of the modified PTH remains unknown, N-PTH appears to be over-secreted in the setting of severe hyperparathyroidism and in parathyroid cancer [34, 35].

Because bioactive assays do not cross-react with large C-PTH fragments, the third-generation-to-second-generation PTH ratio is generally less than one. However, in the setting of parathyroid cancer, this does not appear to be the case. Interestingly, a third-generation-to-second-generation PTH ratio >1 has a sensitivity of 83.3 %, and a specificity of 100 % for diagnosing parathyroid carcinoma in patients presenting with primary hyperparathyroidism [34, 35]. While more studies are needed, this ratio may be a useful adjunct for diagnosing parathyroid carcinoma earlier in the disease course, and identifying those patients who are at higher risk for recurrent disease [34, 35].

Liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS) detection has traditionally been limited to the quantitative analysis of small molecules (e.g., testosterone, vitamin D metabolites). However, MS techniques are rapidly being developed and refined for their ability to quantify larger and more complex molecules, such as PTH_1–84 [36]. To date, the primary challenge preventing widespread implementation of MS for PTH measurement is the significant amount of interference caused by several different modified PTH molecular forms (e.g., oxidized and phosphorylated PTH variants), which can rapidly accumulate in some patient samples. For example, in patients requiring dialysis, a large proportion of PTH becomes oxidized and biologically less active. Thus, results obtained using conventional assays (which include non-oxidized and oxidized PTH) likely over-estimate the true level of biologically active hormone in this patient population.

With this in mind, a variant of the intact PTH assay was recently developed to help improve the specificity of PTH measurements in dialysis patients. First, a monoclonal antibody is used to immobilize and subsequently eliminate all oxidized PTH (identified via methionine residues located at position 8 and/or 18); all remaining PTH is then measured with a conventional intact assay, which employs antibodies against epitopes 26–32 and epitopes 55–64 for capture and detection, respectively [37]. In the original study, a large (but highly variable) proportion (~90 %) of the starting PTH concentration in the dialysis patient samples were determined to be oxidized (biologically less active) [37]. While promising, the clinical implications of differentiating oxidized versus non-oxidized PTH in circulation remain speculative.

Approximately 3–4 mL (minimum 1 mL) of whole blood is collected via venipuncture in a pre-chilled EDTA plasma tube (lavender top). After collection, EDTA whole blood is centrifuged immediately. To avoid hemolysis (which decreases PTH values), the collection tube should not be filled completely, and the sample should remain relatively stable and upright throughout collection, transport, and processing. The specimen is then gently inverted (but NOT shaken) several times to allow mixing. EDTA whole blood is stable for a maximum of 2 h at ambient temperature and 4 h on wet ice.

[Note: Serum (red top) tubes, while a less stable reservoir for PTH compared to EDTA plasma, are also acceptable. However, when using serum tubes, the whole blood sample must be allowed to clot, which can take approximately 30 min (potentially longer in patients receiving anticoagulation therapy) after sample collection. Centrifuging serum samples before clotting occurs can result in fibrin formation. While the time required for whole blood to clot may not be critical for routine clinical or laboratory testing, it can be a source of significant delay in the intraoperative setting (see below).

The plasma (anticoagulated blood) or serum (after a clot has formed) is then separated from the supernatant using a refrigerated centrifuge (10 min at 2095 × g). After centrifugation, EDTA plasma is stable for approximately 4 h at ambient temperature, 24 h at 4 °C, 2–4 months at −20 °C, and 2–24 months at −80 °C [38]. Of note, bioactive (third-generation) assays are associated with greater PTH stability compared to their intact (second-generation) counterparts [39].

Heterophile antibodies in human serum, which may be present in up to 11 % of the population, can potentially bridge the assay immunoglobulins (i.e., the capture and signal antibodies) and falsely elevate in vitro diagnostic results [19, 39, 40]. Heterophile antibodies are becoming increasingly prevalent in the USA because of enhanced exposure to monoclonal antibodies for the treatment of chronic inflammatory disorders, cancer, and transplantation [40]. When heterophile antibodies are suspected, re-analysis of the patient’s serum may be performed using blocking agents that are specific to the immunoassay [40]. Unusual fragments or genetic variations in PTH have been implicated in falsely low measured PTH concentrations for some assays [41].

The International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) established a PTH working group with the intentions of (1) “[defining the] inclusion/exclusion requirements for an appropriate panel of sera and plasma with which to establish reference intervals and establishment of such a panel with support from the clinical community and diagnostic manufacturers,” and (2) “development of a reference measurement procedure for PTH_1–84 to a standard that would enable its adoption by the IFCC reference laboratory network [42].”

Modern or commercially available PTH assays (i.e., intact and bioactive assays) generally yield reference intervals somewhere between 10 and 100 pg/mL. Intact and bioactive assays possess sufficient sensitivity and reproducibility to measure PTH concentrations at the low end of the reference range in the normal population (~10 pg/mL) [43]. When reference subjects with vitamin D levels less than 20 nmol/L (which raises serum PTH levels) are excluded, the upper limit of the normal population PTH reference interval is approximately 50 pg/mL.

Significant variability in results obtained by immunometric assays has repeatedly been demonstrated in the literature [44]. Despite all of the discrepancy, there are no known clinically significant differences in the diagnostic sensitivities between intact (second-generation) and bioactive (third-generation) assays for the diagnosis of primary hyperparathyroidism in patients with normal renal function [45–48]. Moreover, bioactive PTH assays have not yet proved any significant advantage over intact assays in the diagnoses of bone disease or other clinical manifestations of secondary hyperparathyroidism in uremic patients [45].

Assays belonging to the same generation can vary by as much as 4.2-fold in magnitude between the lowest and highest reading methods [44, 49, 50]. The difference in the PTH levels between any two assays, however, is generally proportional throughout the range of measurements. There is some evidence that applying dynamic reference intervals (based on a range of serum PTH values obtained by acute modification of serum calcium concentrations in healthy subjects) instead of Gaussian intervals (based on PTH values observed in individuals with normocalcemic concentrations) may significantly improve the clinical sensitivity of PTH assays ([51]; Lepage 1988). In one study, dynamic reference intervals increased the average clinical sensitivity for detecting primary hyper- and hypoparathyroidism from 68 to 100 % [51].

In 1988, Samuel Nussbaum (Massachusetts General Hospital) and colleagues presented a small case series (N = 13) at the American Association of Endocrine Surgeons (AACE) meeting in Boston suggesting that their PTH immunoradiometric assay (turnaround time 15 min) could be used as an intraoperative adjunct for guiding the extent of neck exploration during parathyroidectomy [27]. In 1993, after Nichols Institute Diagnostics had made available the equipment and labeled antibody for use in an immunochemiluminometric (no longer requiring radioactive isotopes) assay, Dr. George Irvin (University of Miami, Jackson Memorial Hospital) popularized the assay as an adjunct to image-guided, focused parathyroidectomy [52].

Because the plasma half-life of PTH is less than 5 min, intraoperative PTH monitoring enables the surgeon to detect a decline in PTH after the primary source(s) of the excess hormone is excised. The Miami criterion, established by Irvin and continuing to be used today, requires obtaining at least four intraoperative PTH samples (pre-skin incision, pre-gland excision, 5 min post-gland excision, 10 min post-gland excision). A 50 % reduction in PTH levels at 5 or 10 min post-excision is indicative of surgical success. The criteria can be modified or adapted depending on the desired sensitivity and/or specificity [5, 53–55]. For example, many surgeons often also obtain a 20-min sample after gland removal if there is a delayed drop in PTH, which does not meet or is close to the 50 % requirement, before continuing further exploration.

Rapid PTH assays have a turnaround time ranging between 8 and 20 min. Some manufacturers offer a relatively “slower” laboratory mode for routine clinical analysis and a “faster” (e.g., “rapid” or “turbo”) mode for intraoperative analysis. For example, the Access® Immunoassay System (Beckman Coulter, Pasadena, CA) offers “routine” (~30 min) and “intraoperative” (~15 min) modes. According to the manufacturer, when PTH levels are >12 pg/ml, the routine and intraoperative modes’ imprecision levels are <8 % and <12 %, respectively. In our experience, the Beckman intraoperative mode yields about 5 % lower PTH levels compared to the routine mode and the differences in precision were within the manufacturer’s reported ranges (7 % vs. 6 %). See Table 42.1 for other commonly used and commercially available (past and present) intraoperative PTH platforms.

While statistically significant variations among various intraoperative assays have been reported (predominantly in the setting of renal insufficiency), the clinical significance of the variation among commercially available assays is not well established. Technical and patient factors that can cause an insufficient decline in PTH levels include decreased renal clearance in patients with chronic renal disease, significant baseline sample hemolysis or hemodilution (both decrease baseline PTH value), a missed peak, and laboratory errors. Technical factors such as post-excision sample dilution or hemolysis (both decrease PTH value) can lead to a falsely adequate drop in PTH, which in turn can lead to missed parathyroid disease.