World Journal of Surgery

, Volume 34, Issue 3, pp 538–543

Intraoperative Determination of PTH Concentrations in Fine Needle Tissue Aspirates to Identify Parathyroid Tissue During Parathyroidectomy


    • 1st Department of SurgerySemmelweis University
  • László Duffek
    • Department of Radiology and OncotherapySemmelweis University
  • Rezső Szlávik
    • 1st Department of SurgerySemmelweis University
  • István Takács
    • 1st Department of MedicineSemmelweis University
  • Miklós Tóth
    • 2nd Department of MedicineSemmelweis University
  • László RomicsJr.
    • Victoria Infirmary Glasgow

DOI: 10.1007/s00268-009-0351-5

Cite this article as:
Horányi, J., Duffek, L., Szlávik, R. et al. World J Surg (2010) 34: 538. doi:10.1007/s00268-009-0351-5



Exact tissue identification during parathyroidectomy is essential to successfully cure hyperparathyroidism. PTH level determination from tissue aspirates has been advocated as a “biochemical frozen section” for parathyroid tissue identification. We investigated the sensitivity and specificity of this method in a large cohort of consecutive patients who underwent parathyroidectomy in a tertiary referral center.


PTH levels of 359 tissue aspirates were measured intraoperatively in 223 consecutive patients from March 2006 to December 2008. Suspected parathyroid and control tissues were aspirated with a standardized technique immediately after their excision. Samples were processed for quick-PTH assay with peripheral blood samples before and after excision. PTH levels from tissue aspirates were correlated with pathological diagnosis. The Mann–Whitney test was used to determine statistical significance (P < 0.05).


A total of 255 parathyroid (196 adenoma, 30 hyperplasia, 4 carcinoma, 25 normal parathyroid) and 104 nonparathyroid tissue (88 thyroid, 16 lymph node, thymus, or fat) aspirates were compared. A highly significant difference was found between PTH levels of parathyroid (8,120 ± 2,711 pg/ml; interquartile range (IQR): 4,949–9,075) and nonparathyroid (0.8 ± 9.29 pg/ml; IQR: 0.4–1.4) tissue aspirates (P < 0.005). This test is 100% sensitive and 100% specific to identify parathyroid tissue for values >84 pg/ml. Furthermore, PTH levels of pathological parathyroid aspirates (8,169 ± 2,597; IQR: 5,634–9,109) were higher than that of normal parathyroid aspirates (4,130 ± 2,952; IQR: 2,569–8,284; P = 0.0011).


PTH level determination from tissue aspirates is a highly reliable, quick, and simple method to differentiate parathyroid and nonparathyroid tissues during parathyroidectomy. This method can obviate frozen sections in patients undergoing surgery for hyperparathyroidism.


Precise identification of parathyroid tissue is fundamental in parathyroidectomy to correct hyperparathyroidism. Intraoperatively, peripheral venous PTH level determination with quick parathyroid hormone (PTH) assay, radio-guidance with gamma-probe, and frozen-section examination were shown to be reliable means of identifying tissue origin during parathyroid exploration [18]. Recent data suggest, however, that rapid PTH assay is the most useful intraoperative adjunct for patients undergoing surgery for primary hyperparathyroidism [9]. Nevertheless, despite routine application of a combination of pre- and intraoperative localization techniques, as many as 6% of patients with primary hyperparathyroidism will experience persistent disease or will develop a recurrence after initial resection [10].

Certainly, there are numerous conditions in which the sensitivity of the above-mentioned tests is significantly reduced. Previous neck surgery or irradiation can cause an abundance of scar tissue and consequent distortions and adhesions in the anatomical landmarks [11]. Coexisting thyroid disease, such as multinodular goiter or Hashimoto thyroiditis, can make localization and tissue identification much more challenging [11, 12]. Anatomical variations and ectopic parathyroid glands, small or atypical forms of parathyroid glands, or other cervical masses, such as lymph nodes, or multiglandular parathyroid disease can give rise to significant difficulties, as well [1317]. As regards the intraoperative measurement of peripheral venous PTH levels with quick PTH assay, several pitfalls have been described. Involuntary manipulation of the parathyroid gland before the preexcisional blood sample could result in false-positive test, whereas inadvertent manipulation during the removal causes a false-negative test [18, 19]. Furthermore, concomitant thyroid surgery may compromise the blood supply of the parathyroid adenomas, resulting in a misleading decrease of intraoperative PTH levels [20].

Inaccurate diagnoses during frozen-section examination contribute significantly to the operative failures of parathyroidectomy [21, 22]. This is more common in cases of intrathyroidal parathyroid glands, glands with a microfollicular pattern, uncircumscribed oxyphilic cells, or thyroid nodules with fatty stroma [23]. In addition, frozen section can be time-consuming, costly, and requires the excision of a significant portion of tissue, rendering it impractical in many settings. Furthermore, modern minimally invasive parathyroidectomies would be suitable for treatment in ambulatory care hospitals (providing 23-h care), which have gained significant popularity within the United Kingdom recently. Of note, these facilities usually lack on-site pathology department, and therefore, frozen-section examination could not be performed there.

Nevertheless, all the above-discussed methods have significant drawbacks, which would necessitate the development of novel pre- and intraoperative localization and tissue identification techniques. Recently, a few authors following the Perrier et al. pioneering study proposed the intraoperative measurement of intraglandular intact PTH levels from fine needle tissue aspirates to identify parathyroid tissues promptly [1, 1113, 23]. This is a very practical idea, because quick PTH assays have already been used in most centers, and aspirate samples can be run concurrently with the postexcisional peripheral blood sample. This idea originates from the ultrasound-guided preoperative fine-needle aspirations, which is suggested in cases of equivocal ultrasound findings, and the PTH level or cytological analysis of the aspirate predict the presence of parathyroid tissue [15, 16, 2326]. Several authors suggest that intraoperative fine needle aspiration of parathyroid gland is a quick and simple method, and determination of tissue aspirate PTH level with quick PTH assay should replace frozen-section examination during parathyroidectomy [1, 1113, 23]. However, this method can still not be regarded as an established assay. This is due to the very few published studies, relatively low patient numbers in the studies, inconsistency of the aspiration techniques applied, and relatively low case numbers in the control arms as opposed to the study arms.

In our study, therefore, we measured routinely tissue aspirate PTH levels on large number cases with a consistent technique. We completely obviated the use of intraoperative frozen sections during the study period. PTH contents of tissue aspirates were compared with the histological diagnosis of removed putative parathyroid and nonparathyroid tissues to determine the sensitivity of this method.

Patients and methods

PTH levels from tissue aspirations were determined during 223 consecutive operations for hyperparathyroidism in 221 patients (39 men and 182 women; mean age: 61 (range, 20–83) years) from March 2006 through December 2008. Seven patients were referred with recurrent hyperparathyroidism having undergone a previous parathyroidectomy elsewhere. Although a re-do parathyroidectomy was successful in six of seven patients, one patient—who had two previous attempts elsewhere—needed a fourth exploration to achieve a curative resection. Only one patient had a recurrence of which the initial parathyroidectomy was performed by us; her repeat neck exploration was curative, too. Previous thyroidectomies were performed in eight further patients.

Intraoperative determination of PTH levels were performed in 740 parathyroid operations altogether since January 1999, being measured from peripheral venous samples exclusively before March 2006. Patients who were referred with signs, symptoms, and laboratory findings consistent with primary or secondary hyperparathyroidism, as well as with suspicion of multiple endocrine neoplasia (MEN), were all included in this single-arm prospective study. Most of the patients underwent preoperative localizations utilizing ultrasonography and technetium Tc99m sestamibi scintigraphy. A single experienced surgeon performed all aspirations with a standardized technique as well as the operations. There were no surgical trainees or specialist nurses involved in performing the aspirations, because that could have been interfered with the consistency of the technique and influenced the results finally.

Control aspirates for PTH values were taken from thyroid, lymphatic, and fatty tissues. During the 223 parathyroidectomies, 88 patients underwent concomitant partial thyroidectomies for nodular goiter, which was usually revealed by preoperative ultrasonography or as an intraoperative finding.

Suspected parathyroid or control tissues were aspirated immediately after excision to determine their identity, and then the sample was sent for quick PTH assay together with the pre- and postexcisional peripheral blood samples. Tissue aspirations were taken with a standardized technique: a 22-gauge needle (40 mm in length, 0.7 mm in diameter) connected to a 5-ml syringe was punctured into the tissue, moved gently back and forth under suction to aspirate its content into the needle, and removed from the tissue when no suction was applied. The syringes were brought on ice to the core laboratory of the Department of Radiology and Oncotherapy of Semmelweis University. Next, the needle was rinsed with 1 ml of isotonic sodium chloride solution into an appropriate test tube and then vortexed for 1 min. The PTH content of these dilutions was determined similarly to the PTH level of the serum from peripheral blood samples. Intact human PTH was determined with an ultrasensitive immunoradiometric assay (IRMA) using the ELSA-PTH kit (CIS Biointernational; Bioassays, France). This is a solid-phase two-site immunoradiometric kit with a normal range between 10 and 80 pg/ml. The kit showed excellent sensitivity with a detection limit of 6 pg/ml. After 20 min of incubation time, radioactivity was measured with a gamma scintillation counter (Isodata), and concentrations were calculated by means of a standardized curve. ELSA-PTH specifically measures intact human PTH (1-84) with no interference with peptides in this assay.

The values of tissue aspirates were compared with the final pathologic diagnosis; the thyroid gland, lymphatic, and fatty tissues PTH values served as the negative controls. Frozen sections were not performed in any of the above-mentioned cases.

PTH levels were expressed in pg/ml. The Mann–Whitney test was used to compare groups, because none of the groups was normally distributed. Values were expressed as median ± standard deviation, and range and interquartile range (IQR). The difference was statistically significant if P < 0.05.


A total of 359 tissue aspirates were taken and measured for PTH levels: 255 from putative parathyroid glands and 104 from nonparathyroid tissues. Histological analysis revealed that parathyroid aspirates were taken from 230 pathological parathyroid glands and 25 normal parathyroid glands. Of the 230 pathological parathyroid glands, 196 were parathyroid adenoma, 30 were parathyroid hyperplasia, 4 were parathyroid carcinoma, and 25 were normal parathyroid gland. Histological analysis further confirmed that the 104 nonparathyroid tissue aspirates were taken from 88 thyroids (of which incidental papillary microcarcinoma was found in 2 cases and 1 patient with MEN2A syndrome had known medullary carcinoma), 9 lymph nodes, 4 thymus, and 3 fatty tissues.

Aspirates of 255 parathyroid and 104 nonparathyroid tissues were compared in terms of PTH levels and correlated with their pathological diagnosis. Boxplot demonstrates the difference between PTH levels of parathyroid (8,120 ± 2,711; 160–14,530; IQR: 4,949–9,075) and nonparathyroid (0.8 ± 9.29; 0.1–84; IQR: 0.4–1.4) tissues aspirates (Fig. 1). Obviously, the difference between the PTH levels of parathyroid and nonparathyroid tissue aspirates are statistically highly significant (P < 0.005; Mann–Whitney test). In addition to this, for values >84 pg/ml, the test is 100% sensitive and 100% specific to identify parathyroid tissue.
Fig. 1

PTH levels (pg/ml) of parathyroid (8,120 ± 2,711; 160–14,530; IQR: 4,949–9,075) vs. nonparathyroid (0.8 ± 9.29; 0.1–84; IQR: 0.4–1.4) tissues aspirates (P < 0.005; Mann–Whitney test)

Further analysis was performed for 255 parathyroid tissue aspirates, and a comparison was made between the PTH levels of histologically pathological and normal parathyroid tissue aspirates. Boxplot shows the difference between PTH levels of pathological parathyroid (8,169 ± 2,597; 160–14,530; IQR: 5,634–9,109) and normal parathyroid (4,130 ± 2,952; 1,240–9,785; IQR: 2,569–8,284) tissue aspirates (Fig. 2). Statistical analysis revealed that PTH levels of pathological parathyroid tissue aspirates are significantly higher than that of normal parathyroid tissues (P = 0.0011; Mann–Whitney test). Using the 95% confidence interval of histologically confirmed pathological parathyroid glands with the lower limit of a positive test result at 4,989 pg/ml, tissue aspirate PTH measurement has a positive predictive value of 95.3, negative predictive value of 24.6, and 78.7% sensitivity and 64% specificity (true positive: n = 181; true negative: n = 16; false positive: n = 9; false negative: n = 49).
Fig. 2

PTH levels (pg/ml) of pathological parathyroid (8,169 ± 2,597; 160–14,530; IQR: 5,634–9,109) vs. normal parathyroid (4,130 ± 2,952; 1,240–9,785; IQR: 2,569–8,284) tissue aspirates. PTH levels of pathological parathyroid tissue aspirates are significantly higher than that of normal parathyroid tissues (P = 0.0011; Mann–Whitney test)

No significant differences were found between PTH concentrations of aspirates taken from parathyroid adenoma (8,136 ± 2,627; 160–14,530), hyperplasia (8,269 ± 2,465; 1,949–10,951), or carcinoma (8,070 ± 2,968; 2,762–9,370).


Although a few studies in the literature advocate that intraoperative tissue aspirate PTH levels can predict the presence of parathyroid tissue during parathyroidectomies [1, 1113, 23], no large, prospective study has been performed to determine its accuracy. In the previous five studies, the number of patients as well as parathyroid tissue aspirates varied significantly (between 44 and 170) similar to nonparathyroid (control) samples (between 2 and 170). In addition, inconsistency was found in the intraoperative aspiration techniques across the studies, in terms of some applied in vivo, others ex vivo, techniques, whereas different techniques were applied for study and control aspirates occasionally, too. These discrepancies resulted in a rather variable figure for the sensitivity of the method reported, although most authors agreed that this can be used as a biochemical frozen section to confirm the identity of parathyroid tissue. Interestingly, sensitivity ranged from 71% to 100%; the highest figure was published from the smallest study [23]. In this respect, our study certainly has assets in terms of larger sample size and technical consistency.

Since the routine intraoperative PTH assay was introduced at our center in 1999, the overall cure of 517 parathyroidectomies was 97%. A false-positive decrease in the postexcisional PTH level was the most common reason for failure, mostly caused by inadvertent devascularization of the hyperfunctioning parathyroids leading to a decrease even without their removal. This is a well-known phenomenon characterized by a relapse after revascularization of the diseased glands, a common scenario in unrecognized multiglandular disease [20, 27]. During the study period of 223 parathyroidectomies, the overall cure increased to 99.1% when the peripheral blood and tissue aspirate PTH levels were determined concurrently in the same assay intraoperatively.

In general, iPTH values vary from study to study, but our results correlate with the results of most published studies [1, 1113, 23]. Due to the huge differences between the intraoperative aspirate PTH levels of parathyroid and nonparathyroid tissues, most authors assigned a higher cutoff to predict parathyroid tissue identity for practical but not statistical bases, using 1,000 or 1,500 pg/ml. Accepting this rationale our PTH levels from parathyroid aspirates were 2 to 5 decimals higher than that of nonparathyroid tissues. In this sense, of the 255 parathyroid aspirations, there were only 2 aspirates with PTH concentration <1,000 pg/ml (160 and 218 pg/ml from parathyroid adenomas). Similarly, of the 104 nonparathyroid tissue aspirates, there were only three samples with PTH level >10 pg/ml (84 and 22.9 pg/ml from thyroid, and 41 pg/ml from lymphoid tissue). Practically, we can say that parathyroid tissue can be identified if the PTH content is at least 100 times higher than that of controls.

Although most authors agreed that hyperplastic parathyroid glands or parathyroid carcinoma also can be identified with this method [1, 11, 23], it has been debated whether intraoperative determination of tissue aspirate PTH levels could be reliably used to identify normal parathyroid glands. Scepticism regarding normal parathyroids was mainly due to the relatively small size of normal glands and consequent technical difficulties in performing aspirations [1, 23]. Recent studies demonstrated, however, that normal glands can also be reliably determined using this assay, having similar magnitude of difference compared with nonparathyroid tissue aspirates [11, 12]. In our study, we can confirm that tissue aspirate PTH determination is as reliable for hyperplasia and carcinoma as it is for adenoma, because no statistical difference was found in terms of PTH levels in between these three different pathological parathyroid tissue types. Furthermore, PTH levels from normal parathyroid tissue aspirates were 2–5 decimals higher than that of nonparathyroid tissues, similar to pathological parathyroid tissue aspirate PTH levels.

Certain disagreement exists among the previous studies whether intraoperative quick PTH assay from tissue aspirates would be able to differentiate between normal and pathologic parathyroid glands. Concerns arose from differences in tissue size and turgor between normal and pathological parathyroids, making it difficult to sample constant volumes of tissue [12, 23]. However, Kiblut et al. demonstrated that in case of primary hyperparathyroidism, but not in secondary, tissue aspirate PTH levels from parathyroid adenomas are significantly higher than that of normal parathyroids [11]. Importantly, they also showed that PTH values from normal parathyroid glands in patients with primary hyperparathyroidism are lower than levels in patients with thyroid—and not parathyroid—disease only [11]. This could easily be explained by suppression of the normal parathyroid glands in case of primary hyperparathyroidism [26]. Although we aspirated the glands ex vivo, as opposed to in vivo technique used in the previous study, our results can confirm the above-mentioned findings to a certain extent. We could demonstrate some predictivity for pathological parathyroid glands, but not for normal parathyroids. Nevertheless, we believe that these data are insufficient to draw any conclusion at this stage, and therefore, practicality of this result is minor compared with differentiation of parathyroid and nonparathyroid tissue.

Direct aspiration of the suspected parathyroid tissue is a quick and easy technique, although it is operator-dependent, despite the standardization of the technique. Intraoperative determination of tissue aspirate PTH levels is a simple, low-cost test, which can be run simultaneously with the postexcisional peripheral blood PTH assay. Further studies are necessary to determine whether all patients with primary hyperparathyroidism who undergo parathyroidectomy should have PTH measurements from the aspirates of each tissue specimen suspected to be of parathyroid origin, as we did in our study. However, frozen-section examinations were entirely obviated during this series of 223 consecutive parathyroidectomies. We believe, therefore, that tissue aspirate quick PTH assay also can replace frozen-section examination.

Copyright information

© Société Internationale de Chirurgie 2010