• Users Online: 2
  • Print this page
  • Email this page

Table of Contents
Year : 2019  |  Volume : 39  |  Issue : 1  |  Page : 53-59

Intraoperative identification of parathyroid tissue: a comparative validity study of frozen section, cytology, and reflected-light diagnostic methods

1 Department of Pathology, National Cancer Institute, Al-Azhar University, Cairo, Egypt
2 Pathology Department, Faculty of Medicine, Al-Azhar University, Cairo, Egypt

Date of Submission30-Jan-2019
Date of Acceptance02-Feb-2019
Date of Web Publication29-Nov-2019

Correspondence Address:
Nabil El-Bolkainy
Pathology Department, National Cancer Institute, Cairo University, Giza, 11511
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/EGJP.EGJP_8_19

Rights and Permissions

Background Surgery is the main line of treatment for hyperparathyroidism, but preliminary localization of the lesion is essential. This could be accomplished preoperatively by radiographic methods (isotope scans and ultrasonography) or intraoperatively by pathologic methods, namely, frozen section, cytology, and reflected-light microscopy.
Aim To compare the diagnostic accuracy of the intraoperative pathologic methods when used singly or combined. The study was done on 30 (10 parathyroid tissue and 20 nonparathyroid tissue) tissue samples, obtained from nine patients. Modifications were made on a monocular microscope to allow transmitted-light, reflected-light, and digital photography.
Results When used alone, the diagnostic accuracy was 96.6% for frozen section, 86.6% for cytology, and 80% for reflected-light microscopy. The combined use of cytology with reflected-light microscopy increased the diagnostic accuracy to 93.3%, with good concordance with the accuracy of frozen section combined with cytology (κ ratio 0.651). Diagnostic errors were mainly owing to the difficulty to differentiate thyroid from parathyroid tissue.
Conclusion In specialized centers, frozen section combined with cytology is the method of choice for the intraoperative diagnosis of parathyroid tissue. Conversely, in developing countries, where frozen section equipment is usually not available, combination of reflected-light microscopy and cytology is a good, inexpensive, rapid and effective alternative. However, proper training is needed for these methods to ensure an accurate diagnosis.

Keywords: cytology and reflected-light microscopy, frozen section, parathyroid

How to cite this article:
El-Bolkainy T, Rabie A, Zain M, El-Bolkainy N. Intraoperative identification of parathyroid tissue: a comparative validity study of frozen section, cytology, and reflected-light diagnostic methods. Egypt J Pathol 2019;39:53-9

How to cite this URL:
El-Bolkainy T, Rabie A, Zain M, El-Bolkainy N. Intraoperative identification of parathyroid tissue: a comparative validity study of frozen section, cytology, and reflected-light diagnostic methods. Egypt J Pathol [serial online] 2019 [cited 2020 Feb 26];39:53-9. Available from: http://www.xep.eg.net/text.asp?2019/39/1/53/272010

  Introduction Top

Hyperparathyroidism is classified as primary, secondary, and tertiary based on the cause of hormone overproduction. Surgical excision of parathyroid tissue is the main line of treatment, but preliminary localization of the lesion is essential. Primary hyperparathyroidism may be sporadic or syndromic (MEN-1 or MEN IIA associated). It is commonly caused by an adenoma (85%), less commonly by hyperplasia (14%), and rarely by a carcinoma (1%) (Fletcher, 2013). The adenoma is usually solitary and of large size, hence preoperative localization is possible by the combined use of ultrasonography and radiotracer scintigraphy (Johnson, 2010; Vazquez and Richards, 2011). Conversely, secondary hyperparathyroidism and tertiary hyperparathyroidism, which are associated with chronic renal failure, result in multiple gland affection, hence the need of the 4-gland exploration and intraoperative pathologic diagnosis to ensure complete removal of parathyroid tissue.

Three main diagnostic pathologic methods are available to identify parathyroid tissue during operative exploration, namely, frozen section, scrape cytology, and reflected-light microscopy. In specialized centers, frozen section is widely used with an accuracy of 99.2% (Westra et al., 1998). The additional use of cytology increased the diagnostic accuracy of frozen section to 100% (Shidham et al., 2002). Reflected-light microscopy, with toluidine blue staining, offers a rapid inexpensive method of intraoperative tissue diagnosis (El-Bolkainy et al., 1971). However, the application of this method in parathyroid pathology is so far not previously reported. Moreover, it is difficult to differentiate parathyroid adenoma from hyperplasia by all these methods (Lawrence, 1978; Saxe et al., 1985). It requires only a pathologist to distinguish parathyroid tissue from adjacent nonparathyroid tissue such as thyroid tissue, lymph nodes, fat, or thymic tissue.

The aim of this study was to compare the diagnostic accuracy of the aforementioned methods in intraoperative identification of parathyroid tissue. The accuracy was determined for the methods when used singly or in combination. Moreover, an analysis of the main causes of diagnostic errors was made.

  Patients and methods Top


The study was based on nine private patients presenting with hyperparathyroidism treated during the years from 2014 to 2017. Six patients had parathyroid adenomas and three had hyperplasias. An informed consent was obtained from the patients for the operative procedure and the use of tissue for research following the guidelines of the ethical committee of the National Cancer Institute, Cairo University. The final diagnosis was confirmed for all tissue samples by paraffin section histopathology, which serves as a gold standard. During intraoperative exploration of glands, 30 tissue samples were submitted for study to obtain a rapid diagnosis. No parathyroid carcinoma or ectopic thymic tissue was encountered in the present investigation.


The three diagnostic methods were done in a sequential order, namely, first cytology, then reflected-light microscopy, and finally frozen section. This ensured that the diagnosis of cytology and reflected-light microscopy was not influenced by the frozen-section diagnosis. Moreover, this sequence allowed harmony of work between the pathologist and the technician. The former performed the initial two methods, which are rapid, whereas the latter did the final time-consuming frozen section.

For cytology, the scrape-smear method was used, as it is more accurate than the simple imprint method (Rohaizak et al., 2005). The tissue was gently scrapped to allow an adequate sample for the smears which were stained by the rapid hematoxylin and eosin method (Bancroft and Gamble, 2008).The reflected-light microscopy method used is a modification of a previously reported technique (El-Bolkainy et al., 1971). A monocular microscope was equipped by both reflected and transmitted Led-light illumination to allow for both reflected-light and cytologic examinations ([Figure 1]). It was fitted with objective lenses power 4, 10 and 20, as well as a wide field eyepiece (10×20) with camera attachment for Apple iphone-6, 12 megapixel. Fresh tissue was cut (3 mm thick and 2×2 cm size) and placed is a glass  Petri dish More Details (10 cm in diameter). The rapid vital stain 1% toluidine blue was applied by a brush on the upper surface of tissue sample, allowed to stand for 5 s and then washed by water. Before staining, a lubricant gel (K-UY gel, Eco) was applied to the under surface of the tissue to fix it to the dish during washing and also to avoid staining of the under surface of tissue sample by toluidine blue. Finally, the application of the inverted Petri dish cover on the tissue will flatten it to produce a sharp picture ready for examination by both reflected and transmitted light ([Figure 1]).
Figure 1 The monocular combined reflected-transmitted-light microscope (a) camera attachment (b) reflected light. (c) Petri dish containing the stained tissue sample. (d) Transmitted light.

Click here to view

For frozen-section studies, Leica freezing microtomes were used (Leica, Wetzlar, Germany). Tissue sections were stained by the rapid hematoxylin and eosin method with manual agitation (Bancroft and Gamble, 2008).

Statistical methods

The three diagnostic tests, namely, frozen section, cytology, and reflected-light microscopy were first compared regarding their sensitivity, specificity, and accuracy in identifying parathyroid tissue from other tissues (Momeni et al., 2018). Subsequently, the diagnostic accuracy of combined methods was obtained for frozen section and cytology compared with reflected-light and cytology, with concordance analysis using Cohen κ coefficient (Momeni et al., 2018). The causes of diagnostic error were analyzed by identifying the tissues that were misdiagnosed by the different methods. Deferred cases were considered misdiagnosis. For comparison, the final diagnosis made on paraffin sections was used as a gold standard.

  Results Top

The gross appearance of tissue samples was most helpful in diagnosis. Thus, thyroid and parathyroid tissue are tan brown in color ([Figure 2]), whereas lymph nodes were off-white and fatty tissue yellow in color. The size of normal parathyroid gland varied between 3 and 6 mm in longest diameter, whereas adenomas and hyperplastic glands are much larger, ranging from 1 to 4 cm.
Figure 2 Gross pathology of parathyroid adenoma, a large solitary tan brown tumor.

Click here to view

The cytomorphologic features of different samples are presented in [Figure 3]. Parathyroid tissue was characterized by small cells with scanty cytoplasm, occurring separately or in clusters ([Figure 3]a); follicular pattern was rarely observed ([Figure 3]b). Conversely, thyroid tissue commonly showed follicular pattern ([Figure 3]c), as well as background rich in colloid and histiocytes ([Figure 3]d). In lymph node samples, dissociated lymphocytes with indistinct cytoplasm were a diagnostic feature ([Figure 3]e). Soft-tissue fat appeared as large cells with eccentric nuclei and large cytoplasmic vacuoles ([Figure 3]f).
Figure 3 Cytomorphology, hematoxylin and eosin stain, ×200. (a) Parathyroid smear showing small round and oval cells in clusters related to blood vessels. (b) Most cells have cytoplasm, but few have naked nuclei or large nuclei; chromatin is fine and dispersed, with unusual follicular pattern. (c) Thyroid hyperplasia with numerous follicular pattern. (d) Thyroid hyperplasia with background rich in colloid with crack artifacts and few histiocytes. (e) Lymph node smears showing dispersed lymphocytes with indistinct cytoplasm and clumped chromatin. (f) Fatty tissue smear showing large cells with cytoplasmic vacuoles and eccentric nuclei with associated blood vessels.

Click here to view

Reflected-light pictures of tissue samples are shown in [Figure 4]. Parathyroid tissue appeared as small cells with solid pattern, lacking follicular or nodular features, but few fat cells may be observed, especially in normal parathyroid gland ([Figure 4]a). Thyroid tissue showed a prominent follicular pattern ([Figure 4]b), whereas lymph nodes were identified by the presence of germinal centers ([Figure 4]c), and fatty tissue showed the natural yellow color of fat ([Figure 4]d).
Figure 4 Reflected-light histomorphology, toluidine blue stain, ×40 (a) Normal parathyroid tissue showing small cells related to blood vessels. (b) Thyroid tissue showing numerous follicles. (c) Lymph node tissue with multiple germinal centers. (d) Adipose tissue with diagnostic natural yellow color of fat.

Click here to view

In normal parathyroid gland, fat cells were commonly observed ([Figure 5]a). The histomorphology of adenoma and hyperplasia is basically similar, with presence of three cell types, namely, chief, oxyphil, and clear cells ([Figure 5]b). Focal nodular pattern was observed in hyperplastic glands ([Figure 5]c). In thyroid tissue, a prominent follicular pattern with colloid is diagnostic ([Figure 5]d). Lymph nodes were identified by presence of germinal centers ([Figure 5]e), and adipose tissue by the characteristic fat cells ([Figure 5]f).
Figure 5 Frozen section histopathology, hematoxylin and eosin, ×100. (a) Normal parathyroid gland, small size (<5 mm) showing vascular stroma with few scattered fat cells. (b) Parathyroid adenoma with three cell types, namely, chief cells (right field), oxyphil cells (left field), and clear cells (upper field), arranged in a vascular and trabecular pattern. (c) Secondary parathyroid hyperplasia with focal nodular pattern. (d) Thyroid tissue showing numerous follicles filled with colloid. (e) Lymph node with prominent germinal centers. (f) Fibroadipose tissue, characterized by fat cells with large cytoplasmic vacuoles and fibrovascular septa.

Click here to view

A comparison of the diagnostic accuracy of the methods, as well as their diagnostic errors when used alone is presented in [Table 1]. Frozen section was the most accurate (96.6%), whereas reflected-light microscopy was the least accurate (80%). Only three tissues were misdiagnosed, including one case of misjudgment and two deferred cases that were counted as negative results. [Table 2] shows discordance of cytology and reflected-light microscopy when used alone as compared with frozen section, with a fair κ ratio of 0.366 and 0.242, respectively. However, when reflected-light microscopy was used in combination with cytology, an increase in accuracy was obtained (accuracy of 93.3%). [Table 3] with a good concordance (κ ratio of 0.651). Regarding the time factor, cytologic and reflected-light studies were completed in 4 min, but frozen section took 20 min.
Table 1 Comparison of diagnostic accuracy of pathologic methods and causes of error (n=30)

Click here to view
Table 2 Discordance of cytology and reflected-light microscopy when used alone as compared with frozen section (n=30)

Click here to view
Table 3 Concordance of diagnostic accuracy of reflected-light microscopy and frozen section when combined with cytology (n=30)

Click here to view

  Discussion Top

Historically, frozen section was first used in 1891 by Welsh and Halsted to confirm the diagnosis of breast cancer before mastectomy. Terry in 1920 introduced his rapid razor section method with methylene blue staining as an alternative to frozen section. However, this method failed to gain popularity. Imprint cytology was introduced by Dudgeon and Patrick in 1927 and proved to be a valuable intraoperative diagnostic tool. Winkel and Zeiss in 1927 invented the first reflected-light microscope for microdissection purposes. The subsequent advanced models were used mainly in industrial rather than diagnostic medical applications. In case of intraoperative diagnosis of parathyroid tissue, there are several previous reports on the use of frozen section and cytology (Shidham et al., 2002; Anton and Wheeler, 2005; Rohaizak et al., 2005; Johnson, 2010; Warpe and Agale, 2015), but to the best of our knowledge, no previous studies exist on reflected-light microscopy.

When used alone, the diagnostic accuracy in this study was 96.6% for frozen section, 86.6% for cytology, and 80.0% for reflected-light microscopy. A higher accuracy rate for cytology of 98.0% was previously reported (Shidham et al., 2002). However, the diagnostic accuracy of frozen section in this study is in agreement with the previous reports of 94% (Shidham et al., 2002) and 99.2% (Westra et al., 1998). The combined use of cytology with reflected-light microscopy increased the diagnostic accuracy of the latter from 80.0 to 93.3%, an accuracy figure in good concordance with frozen section. Such an increase in accuracy is explained by the combination of advantages of each method when used together. Thus, reflected-light microscopy is efficient in recognizing tissue pattern, whereas cytology is ideal to identify cellular details. In developing countries where frozen section equipment are unavailable in many hospitals, the combined use of reflected-light and cytology is an efficient alternative diagnostic method. The few errors that may occur are the result of both technical and interpretation problems (Yao et al., 2003).

It is important to differentiate primary from secondary and tertiary hyperparathyroidism, as surgical treatment is different. Primary hyperparathyroidism is usually caused by a solitary adenoma, associated with normal other parathyroid glands. Conversely, secondary and tertiary hyperparathyroidism present with multiple gland affection by hyperplasia. It is impossible to differentiate between and adenoma and hyperplastic gland by histology (Lawrence, 1978; Saxe et al., 1985). Parathyroid hyperplasia is usually associated with chronic renal disease and requires near-total excision of all affected glands. Conversely, in case of adenoma, preoperative localization of the tumor is possible by scans (Johnson, 2010; Vazquez and Richards, 2011), and surgical excision of the solitary tumor only is sufficient. However, confirmation of diagnosis is needed by intraoperative exploration to confirm the presence of at least one normal parathyroid gland (Wick et al., 2015). In addition, confirmation of complete excision of parathyroid tissue is possible intraoperatively by parathormone assays (Irvin and Deriso, 1994). As the half-life of parathormone is only 2–5 min, completeness of resection is confirmed by the decrease of serum parathormone levels 10 min after excision.

  Conclusion Top

In specialized centers, frozen section combined with cytology is the method of choice for the intraoperative diagnosis of parathyroid tissue. However, where frozen section equipment is not available, combination of reflected-light microscopy and cytology is an inexpressive, rapid, and effective alternative. However, proper training is needed in these methods to ensure an accurate diagnosis.[16]

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

  References Top

Anton RC, Wheeler TM (2005). Frozen section of thyroid and parathyroid specimens. Arch Pathol Lab Med 139:1575–1584.  Back to cited text no. 1
Bancroft JD, Gamble MG (2008). Theory and practice of histological techniques. 6th edition. Philadelphia, PA: Churchill Livingstone, Elsevier. p. 127.  Back to cited text no. 2
El-Bolkainy MN, Solimen OMA, Mansour MA (1971). Rapid diagnosis of breast lesions by reflected-light microscopy. Med J Cairo Univ 39:1–10.  Back to cited text no. 3
Fletcher CDM (2013). Diagnostic histopathology of tumors. 4th ed. Philadelpia, PA, USA: El-Sevier, Saunders. p. 1273.  Back to cited text no. 4
Irvin GI, Deriso GT (1994). A new practical intraoperative parathyroid hormone assay. Am J Surg 168:466–468.  Back to cited text no. 5
Johnson SI (2010). Changing clinicopathological practice in parathyroid disease. Histopathol 56:835–851.  Back to cited text no. 6
Lawrence DA (1978). A histological comparison of adenomatous and hyperplastic parathyroid glands. J Clin Pathol 31:626–632.  Back to cited text no. 7
Momeni A, Pincus M, Libien J (2018). Introduction to statistical methods in pathology. Basel, Switzerland: Springer International Publishing AG. p. 18  Back to cited text no. 8
Rohaizak M, Munchar MJ, Meah FA, Jasmi AY, Julia Munchar MJ (2005). Prospective study comparing scrape cytology with frozen section cytology with frozen section is the intraoperative identification of parathyroid tissue. Asian J Surg 28:82–85.  Back to cited text no. 9
Saxe AW, Baier R, Tesluk H, Toreson W (1985). The role of the pathologist in the surgical treatment of hyperparathyroidism. Surg Gynecol Obstet 16:101–105.  Back to cited text no. 10
Shidham VB, Asma Z, Rao RN, Chavan A, Machhi J, Almagro U, Komorowski RA et al. (2002). Intraoperative cytology increases the diagnostic accuracy of frozen sections for the confirmation of various tissues in the parathyroid region. Am J Clin Pathol 18:895–902.  Back to cited text no. 11
Vazquez BJ, Richards MI (2011). Imaging of the thyroid and parathyroid glands. Surg Clin North Am 91:15–32.  Back to cited text no. 12
Warpe BM, Agale SV (2015). Diagnosis of parathyroid adenoma using intraoperative squach cytology and frozen section: a rare case report. IOSR J Dent Med Sci 14:56–58.  Back to cited text no. 13
Westra WH, Pritchett DD, Udelsman R (1998). Intraoperative confirmation of parathyroid tissue during parathyroid exploration: a retrospective evaluation of frozen section. Am J Surg Pathol 22:538–544.  Back to cited text no. 14
Wick MR, Livolci VA, Pfeifer JD, editors (2015). Silverberg principles and practice of surgical pathology and cytopathology. 5th ed. Cambridge University Press. pp. 2949–2975.  Back to cited text no. 15
Yao DX, Hoda SA, Yin DY, Kuhel WI, Harigopal M, Resetkova E, DeLellis RA (2003). Interpretation problems and preparative technique influence reliability of intraoperative parathyroid touch imprints. Arch Pathol Lab Med 127:64–67.  Back to cited text no. 16


  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]

  [Table 1], [Table 2], [Table 3]


    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Access Statistics
    Email Alert *
    Add to My List *
* Registration required (free)  

  In this article
Patients and methods
Article Figures
Article Tables

 Article Access Statistics
    PDF Downloaded12    
    Comments [Add]    

Recommend this journal