Original Article

Cytologic Features of Parathyroid Fine-Needle Aspiration on ThinPrep Preparations Shelley I. Odronic, MD; Jordan P. Reynolds, MD; and Deborah J. Chute, MD

BACKGROUND: Previous studies have provided cytologic criteria that aid in the recognition of parathyroid tissue on aspirate smears, including high cellularity, the presence of naked nuclei, loose 2-dimensional clusters, and papillary architecture. However, to the authors knowledge, the cytomorphologic features of parathyroid fine-needle aspiration (FNA) on liquid-based preparations have not been previously described. METHODS: The authors retrospectively reviewed all parathyroid FNAs that had aspirate smears and a ThinPrep preparation performed over 10 years at 1 institution. The FNA smears and ThinPrep preparations from each case were deidentified and independently reviewed for cellularity, naked nuclei, architecture, and colloid-like material. RESULTS: Forty patients were included in the current study. When individual cases were compared, the ThinPrep preparation was more likely to have lower cellularity, lack papillary architecture, lack naked nuclei, and have areas with a microfollicular pattern compared with the FNA smear. CONCLUSIONS: The cytologic features of parathyroid tissue vary depending on preparation. Many of the common features of parathyroid aspirates are lost on ThinPrep preparations, and an increased percentage of parathyroid FNA specimens have a microfollicular pattern on ThinPrep. This may lead to difficulty in recognizing parathyroid origin on FNA. Cancer (Cancer Cytopathol) C 2014 American Cancer Society. 2014;122:678-84. V

KEY WORDS: biopsy; endocrine; head and neck; liquid-based preparation; nongynecologic cytology; thyroid.

INTRODUCTION Fine-needle aspiration (FNA) is a well-established diagnostic method for thyroid lesions. Traditionally, thyroid FNA has relied on conventional smears; however, liquid-based preparations are increasingly being used for nongynecologic specimens in cytology.1 Recently, there have been several studies comparing the cytomorphologic features of conventional smears with liquid-based preparations for thyroid FNA.2 Advantages of liquid-based preparations include the concentration of material on a slide resulting in decreased screening time, slide quality that is independent of the level of experience of the individual making the smear, and the ability to make additional slides or cell block for molecular tests and immunohistochemistry.1,3 In addition, the liquid-based preparations have less air-drying artifact, are less obscured by blood, and may have a lower rate of unsatisfactory specimens.2-4 In fact, some institutions only use liquid-based preparations in the diagnosis of thyroid lesions by FNA.5-7 Previous studies have established that cytomorphologic features in thyroid FNAs differ between liquidbased preparations and conventional smears.8,9 A common pitfall in the diagnosis of thyroid nodules is the inadvertent sampling of parathyroid tissue, which may be difficult to distinguish from thyroid tissue on FNA due to similar cytologic features.10,11 Previous studies have characterized the cytomorphology of parathyroid gland tissue on aspirate smears,10-16 which include the following: high cellularity, the presence of naked nuclei, loose 2-dimensional clusters, and papillary Corresponding author: Deborah J. Chute, MD, Pathology and Laboratory Medicine Institute, Cleveland Clinic, 9500 Euclid Ave, Cleveland, OH 44195; Fax: (216) 445-3707; [email protected] Pathology and Laboratory Medicine Institute, Cleveland Clinic, Cleveland, Ohio Received: March 20, 2014; Revised: May 7, 2014; Accepted: May 27, 2014 Published online June 18, 2014 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/cncy.21453, wileyonlinelibrary.com


Cancer Cytopathology

September 2014

Cytology of Parathyroid FNA on ThinPrep/Odronic et al

architecture on FNA smears that may be suggestive of parathyroid tissue. With the increasing use of liquidbased preparations for thyroid FNA, there is a need to characterize the findings of parathyroid tissue on liquidbased preparations, which to our knowledge has not been previously described. The current study aimed to characterize the cytologic features of parathyroid tissue on ThinPrep preparations.

MATERIALS AND METHODS All patients with a parathyroid FNA and an associated surgical resection performed over a 10-year period (20022012) at 1 institution were included in the current study. Parathyroid tissue origin was confirmed in all cases by an elevated FNA parathyroid hormone (PTH) level in the needle rinse compared with the serum PTH. All aspirates were performed by a clinician under ultrasound guidance, and the number of passes was at the discretion of the clinician. Papanicolaou-stained smears were prepared from the sample. The ThinPrep slide (Hologic Inc, Bedford, Mass) was prepared according to the manufacturer’s instructions from needle rinse material. Cases were excluded if a ThinPrep slide was not available for review. The FNA smears and ThinPrep preparation from each case were separated and independently reviewed by 2 pathologists blinded to the appearance of other preparations. The FNA smears and ThinPrep preparation were assessed for the following features of parathyroid tissue previously reported in aspirate smears14: cellularity (low [500 cells per slide]), the presence of naked nuclei, architectural patterns present (loose 2-dimensional clusters, papillary structures, and/or microfollicles), and the presence of colloid-like material. The predominant architectural pattern for each case and preparation type was also recorded. Discrepant results were reviewed using a dual-headed microscope and a consensus was reached. A Fisher exact test was used to perform all statistical analysis.

RESULTS Forty patients underwent parathyroid FNA during the study period and were included in the current study (13 men and 27 women). The mean patient age was 58 years (range, 32 years-79 years). The clinical indication for FNA was hyperparathyroidism in all patients (33 patients Cancer Cytopathology

September 2014

with primary hyperparathyroidism, 4 patients with secondary hyperparathyroidism, 2 patients with multiple endocrine neoplasia-related hyperparathyroidism, and 1 patient with a history of parathyroid carcinoma). The preFNA mean serum PTH level was 210 pg/mL (range, 58 pg/mL-542 pg/mL). The site of the FNA was specified as parathyroid for 23 patients, thyroid for 11 patients, neck mass for 3 patients, left cervical nodule for 1 patient, left lower neck for 1 patient, and mediastinal mass for 1 patient. The median number of FNA slides made per case was 4 (range, 3-9 slides per case). The final cytologic interpretation was cytologically bland epithelial cells consistent with parathyroid tissue in 37 cases and atypical cells consistent with parathyroid tissue in 3 cases. The final diagnosis noted high cellularity in 24 cases and low cellularity in 2 cases; cellularity was not mentioned in the final report in 14 cases. Two cases mentioned changes consistent with a cyst. On the surgical resection specimen, the final histologic diagnosis was hypercellular parathyroid gland in 34 patients, intrathyroidal hypercellular parathyroid in 3 patients, parathyroid adenoma in 2 patients, and recurrent parathyroid carcinoma in 1 patient. Measurement of Serum PTH Levels

The mean aspirate PTH level was 44,832 pg/mL (range, 506 pg/mL-397,500 pg/mL). The average ratio of FNA PTH to serum PTH was 270 (range, 2.5-1458). No cases had immunohistochemistry for PTH performed on the cytology specimen. Cytologic Findings

Parathyroid cells on ThinPrep preparations demonstrated round, centrally placed nuclei with stippled nuclear chromatin (Fig. 1, upper left). The nuclei often appeared smaller and darker on ThinPrep preparations than what was observed on the corresponding FNA smears. The cytoplasm of parathyroid cells on ThinPrep preparations was moderate to scant, and had indistinct cytoplasmic borders. The cytoplasm was predominantly lacy, although in some cases with oncocytic differentiation the cytoplasm had a granular appearance with increased density (Fig. 1, upper right). Architecturally, the parathyroid cells were usually loosely clustered into small groups, often with a syncytial arrangement forming branching, loose, 2dimensional clusters (Fig. 1, lower left) or were present as single cells. The background was generally clean, but 679

Original Article

Figure 1. Parathyroid fine-needle aspiration samples on ThinPrep preparation are shown demonstrating small, round, dark nuclei with (Upper Left) indistinct lacy cytoplasm (Papanicolaou stain, 3 400), (Upper Right) oncocytic changes with abundant granular cytoplasm (Papanicolaou stain, 3 400), (Lower Left) loose 2-dimensional clusters in a clean background lacking naked nuclei (Papanicolaou stain, 3 200), and (Lower Right) a background of histiocytes and colloid-like material (Papanicolaou stain, 3 200).

some cases contained hemosiderin-laden macrophages and colloid-like material (Fig. 1, lower right). Naked nuclei were distinctly uncommon on ThinPrep preparations. There was no mitotic activity, significant pleomorphism, or prominent nucleoli identified in any case, including the case of recurrent parathyroid carcinoma. The cytologic differences between FNA smears and ThinPrep preparations are shown in Table 1. Approximately 65% of FNA smears showed high cellularity and 10% demonstrated low cellularity. In contrast, only 40% of ThinPrep preparations had high cellularity and 37.5% had low cellularity. Naked nuclei were present in 85% of FNA smears compared with 57.5% of ThinPrep preparations. Colloid-like material was present in 12.5% of both FNA smears and ThinPrep preparations. All architectural patterns present were noted for each case on the FNA smear and ThinPrep preparation, as well as the predominant architectural pattern present. Papillary structures were common on FNA smears (50%), but were found to be present on only 1 ThinPrep 680

preparation (4%). Loose 2-dimensional clusters were present in almost all FNA smears (95%) and ThinPrep preparations (93%). Microfollicular architecture was present in 76% of FNA smears, but was more common on ThinPrep preparations (96%). The predominant architectural pattern on FNA smears was most commonly loose 2-dimensional clusters (61%) followed by papillary groups (26%). In contrast, the predominant pattern on ThinPrep preparations was most commonly loose 2-dimensional clusters (70%) followed by a microfollicular arrangement of cells (30%). When individual cases were compared, the ThinPrep preparation was more likely to demonstrate lower cellularity, lack papillary architecture, lack naked nuclei, and have areas with a microfollicular pattern (Figs. 2-4).

DISCUSSION Liquid-based preparations are increasingly used in thyroid FNA interpretation, and they are used in some institutions without conventional smears.2,5,7,17 Therefore, it is Cancer Cytopathology

September 2014

Cytology of Parathyroid FNA on ThinPrep/Odronic et al

important to characterize the cytomorphology of possible pitfalls, such as parathyroid sampling, in liquid-based preparations. In a recent study conducted by the College of American Pathologists that investigated the diagnostic accuracy of thyroid FNA liquid-based preparations versus conventional smears, the authors attributed the reduced ability of participants to diagnose papillary thyroid carcinoma on liquid-based preparations to the lack of familiarity with its appearance on liquid-based preparations.2 Compared with FNA smears, liquid-based preparations

TABLE 1. Cytologic Features of Parathyroid on FNA Smears Versus ThinPrep Preparations (n540) Characteristic Cellularity Low Intermediate High Total Predominant architectural patternb Papillary Loose 2-dimensional clusters Microfollicular Architectural patterns presentc Papillary Loose 2-dimensional clusters Microfollicular Naked nuclei Colloid-like material

FNA Smear No. (%)

Thin Prep No. (%)

4 (10) 10 (25) 26 (65) 40

15 (37.5) 9 (22.5) 16 (40) 40

10 (26) 23 (61) 5 (13)

0 (0) 19 (70) 8 (30)

19 (50) 36 (95) 29 (76) 34 (85) 5 (12.5)

1 (4) 25 (93) 26 (96) 23 (57.5) 5 (12.5)




NS .026a NS

Abbreviations: FNA, fine-needle aspiration; NS, not significant. a Determined using the Fisher exact test. b Cases with limited cells for evaluation were excluded from the assessment of architectural patterns. c Cases frequently had >1 architectural pattern present, whereas cases with limited cells for evaluation were excluded from the assessment of architectural patterns.

of thyroid aspirates typically show less colloid, cell shrinkage, and greater nuclear detail.3,18,19 Inadvertent or deliberate sampling of parathyroid tissue can be a significant pitfall in thyroid FNA, particularly if the patient’s history of hyperparathyroidism is not provided. Several previous studies have highlighted the importance of recognizing the cytologic features of parathyroid sampling to prevent misdiagnosis.10-12,14,15 Parathyroid tissue on FNA can be difficult to identify due to overlapping features with thyroid neoplasms, particularly follicular neoplasms.14,15 Parathyroid tissue has been misdiagnosed as follicular thyroid carcinoma,10,13,15 papillary thyroid carcinoma,13 and medullary thyroid carcinoma on FNA.11 The sensitivity of FNA in the diagnosis of parathyroid tissue has been reported to be between 40% and 86%.12,16,20,21 The previous literature regarding FNA of the parathyroid gland has only characterized the cytomorphology on FNA smears. Characteristics such as high cellularity, the presence of naked nuclei, loose 2-dimensional clusters, and papillary architecture on FNA smears are suggestive of parathyroid tissue.12,15 To our knowledge, the current study is the first to examine the cytologic features of parathyroid tissue on liquid-based preparations and how they differ from FNA smear preparations. The current study found that the most common cytologic features of parathyroid tissue differ between ThinPrep preparations and FNA smears. ThinPrep preparations rarely demonstrate papillary architecture and often show a microfollicular pattern. Only 1 case showed papillary architecture on the ThinPrep preparation and it was not the predominant pattern. The prevalence of a

Figure 2. Comparison of parathyroid fine-needle aspiration (FNA) smear and ThinPrep preparation appearance is shown for case 24. (Left) The FNA smear demonstrates papillary architecture, high cellularity, and naked nuclei (Papanicolaou stain, 3100), whereas (Right) the ThinPrep preparation has a similar papillary architecture, but a loss of background naked nuclei (Papanicolaou stain, 3200).

Cancer Cytopathology

September 2014


Original Article

Figure 3. Comparison of parathyroid fine-needle aspiration (FNA) smear and ThinPrep preparation appearance is shown for case 8. (Left) The FNA smear shows loose 2-dimensional groups and numerous background naked nuclei (Papanicolaou stain, 3200), whereas (Right) the ThinPrep preparation has lower cellularity, histiocytes, and colloid-like material, mimicking thyroid material (Papanicolaou stain, 3200).

Figure 4. Comparison of parathyroid fine-needle aspiration (FNA) smear and ThinPrep preparation appearance is shown for case 38. (Left) The FNA smear shows loose 2-dimensional clusters, high cellularity, and rare naked nuclei (Papanicolaou stain, 3100), whereas (Right) the ThinPrep preparation has substantially lower cellularity, smaller epithelial fragments, and a microfollicular pattern (Papanicolaou stain, 3200).

microfollicular architectural pattern on ThinPrep preparations (present in 96% of cases) could potentially contribute to the overinterpretation of parathyroid tissue as “suspicious for follicular neoplasm” of the thyroid on FNA. In addition, the ThinPrep preparations contain less overall cellularity and fewer naked nuclei. The naked nuclei were also difficult to distinguish from background inflammatory cells. It is unclear why there are differences in parathyroid cytology between FNA smears and ThinPrep preparations, but it may be due to the dispersion of cell clusters and papillary structures during ThinPrep processing, or a more limited sample present in the needle rinse after FNA smears are made. Despite the differences between preparations, the presence of loose 2-dimensional clusters and naked nuclei 682

continue to be consistent features in both FNA smears and ThinPrep preparations. Some studies have noted an absence of colloid as a helpful feature in parathyroid aspirates.16,20 However, colloid-like material can be produced by hyperplastic parathyroid glands and true colloid can be acquired as the needle crosses thyroid tissue en route to parathyroid tissue. In the current study, 12.5% of cases contained colloid-like material on both FNA smears and ThinPrep preparations. This is consistent with previous studies, which have found colloid-like material in 8% to 24% of parathyroid FNAs.12-16,21 Similarly, contaminating thyroid epithelial cells have been reported in 8% to 31% of parathyroid FNAs.12,20,21 The case of parathyroid carcinoma in the current study demonstrated cytologic features similar to Cancer Cytopathology

September 2014

Cytology of Parathyroid FNA on ThinPrep/Odronic et al

parathyroid hyperplasia on FNA smear or ThinPrep preparation. Atypia was not appreciated, possibly due to limited cellularity. Previous studies have found that parathyroid carcinomas can have bland cytology, rendering them indistinguishable from otherwise benign parathyroid lesions,22 because histologic confirmation of invasion is required in most cases for the diagnosis of parathyroid carcinoma. Similarly, the current study found no difference between the cytologic features of parathyroid hyperplasia and parathyroid adenoma, although it only included 2 cases of parathyroid adenoma. This is similar to the findings of previous studies,21 although some authors have suggested that there are subtle differences on cytology.23 Ancillary studies, including FNA PTH level and immunohistochemistry, are very helpful in the diagnosis of parathyroid sampling on cytology. FNA PTH levels are highly sensitive and specific in the identification of parathyroid tissue sampling on FNA.24 Owens et al highlighted how the FNA PTH level is helpful to appropriately triage patients when there is suspected sampling of parathyroid tissue.11 Most studies recommend a ratio of FNA PTH level to serum PTH level of >1 for the identification of parathyroid tissue.11,24,25 This ratio was measured in all of the cases in the current study, with a mean FNA-PTH-to-serum-PTH ratio of 270. Unfortunately, the physician performing the FNA must send a sample for this assay at the time of FNA because it requires a fresh sample in normal saline or a similar balanced salt solution.11,24,25 A FNA PTH test cannot be performed retrospectively on fixed needle rinse material in CytoLyt (Cytyc Corporation, Marlborough, Mass)or PreservCyt (Hologic Inc). Immunohistochemistry for PTH may also be helpful if there is adequate material,14 particularly if a FNA PTH level was not performed at the time of FNA. An optimal panel includes PTH and thyroglobulin stains.26 This is due to the tendency for some parathyroid lesions to be falsely negative for PTH immunohistochemical staining if they lack sufficient hormone in individual cells.15 Therefore, thyroglobulin staining provided a control with which to confirm the absence of thyroid tissue. Unfortunately, in cases with a hypocellular cell block or laboratories that lack the capability to perform immunohistochemistry on cytology slides, morphologic features alone may be the only available modality for diagnosing these lesions. Cancer Cytopathology

September 2014

CONCLUSIONS In most situations, a clinical history of hyperparathyroidism, an elevated FNA PTH level, or an indication from the clinician that the FNA is likely from a parathyroid gland will help direct the pathologist to the correct diagnosis. However, knowledge of the cytomorphologic features of parathyroid gland FNA can be helpful in confirming the clinical impression, and may be critical to making the correct diagnosis in cases in which suspicion of parathyroid sampling is not provided or considered by the clinician. The results of the current study indicate that parathyroid cytomorphology differs between ThinPrep preparations and FNA smears. Common features suggestive of parathyroid tissue, such as papillary architecture, naked nuclei, and high cellularity, may not be present or as prominent on ThinPrep preparations. Furthermore, ThinPrep preparations more commonly demonstrate a microfollicular pattern, which may compound the difficulty in distinguishing parathyroid tissue from a thyroid follicular neoplasm. Although the current study did not directly address the sensitivity of any specific cytologic feature on ThinPrep preparations to differentiate parathyroid sampling from thyroid epithelium, it highlights the increasing potential for misinterpreting parathyroid epithelium as thyroid sampling on ThinPrep preparations. Review of clinical and radiologic information and correlation with ancillary studies may be helpful in cases in which there is concern for parathyroid gland aspiration. FUNDING SUPPORT No specific funding was disclosed.

CONFLICT OF INTEREST DISCLOSURES The authors made no disclosures.





Stamataki M, Anninos D, Brountzos E, et al. The role of liquidbased cytology in the investigation of thyroid lesions. Cytopathology. 2008;19:11-18. Fischer AH, Clayton AC, Bentz JS, et al. Performance differences between conventional smears and liquid-based preparations of thyroid fine-needle aspiration samples: analysis of 47,076 responses in the College of American Pathologists Interlaboratory Comparison Program in Non-Gynecologic Cytology. Arch Pathol Lab Med. 2013;137:26-31. Malle D, Valeri RM, Pazaitou-Panajiotou K, Kiziridou A, Vainas I, Destouni C. Use of a thin-layer technique in thyroid fine needle aspiration. Acta Cytol. 2006;50:23-27. Fadda G, Rossi ED. Liquid-based cytology in fine needle aspiration biopsies of the thyroid gland. Acta Cytol. 2011;55:389-400.


Original Article 5. 6.











Ljung BM. Thyroid fine-needle aspiration: smears versus liquidbased preparations. Cancer. 2008;114:144-148. Chang H, Lee E, Lee H, Choi J, Kim A, Kim BH. Comparison of diagnostic values of thyroid aspiration samples using liquid-based preparation and conventional smear: one-year experience in a single institution. APMIS. 2012;121:139-145. Yassa L, Cibas ES, Benson CB, et al. Long-term assessment of a multidisciplinary approach to thyroid nodule diagnostic evaluation. Cancer. 2007;111:508-516. Lee JS, Choi HS, Park IA, Ryu HS. Liquid-based fine needle aspiration biopsy of papillary thyroid carcinoma: logistic regression analysis with conventional and new cytomorphologic features. Acta Cytol. 2013;57:233-240. Tetikkurt US, Oz Puyan F, Oz F, Erdogan N, Ceylan S, Yakupoglu A. Diagnostic value of liquid-based (Liqui-PREP) preparations and interobserver reproducibility in fine needle aspiration cytology of the nodular thyroid lesions. Diagn Cytopathol. 2012;40: 388-393. Layfield LJ. Fine needle aspiration cytology of cystic parathyroid lesions. A cytomorphologic overlap with cystic lesions of the thyroid. Acta Cytol. 1991;35:447-450. Owens CL, Rekhtman N, Sokoll L, Ali SZ. Parathyroid hormone assay in fine-needle aspirate is useful in differentiating inadvertently sampled parathyroid tissue from thyroid lesions. Diagn Cytopathol. 2008;36:227-231. Agarwal AM, Bentz JS, Hungerford R, Abraham D. Parathyroid fine-needle aspiration cytology in the evaluation of parathyroid adenoma: cytologic findings from 53 patients. Diagn Cytopathol. 2009; 37:407-410. Bondeson L, Bondeson AG, Nissborg A, Thompson NW. Cytopathological variables in parathyroid lesions: a study based on 1,600 cases of hyperparathyroidism. Diagn Cytopathol. 1997;16:476-482. Dimashkieh H, Krishnamurthy S. Ultrasound guided fine needle aspiration biopsy of parathyroid gland and lesions. Cytojournal. 2006;3:6. Tseleni-Balafouta S, Gakiopoulou H, Kavantzas N, Agrogiannis G, Givalos N, Patsouris E. Parathyroid proliferations: a source








23. 24.



of diagnostic pitfalls in FNA of thyroid. Cancer. 2007;111:130136. Kini U, Shariff S, Thomas JA. Ultrasonically guided fine needle aspiration of the parathyroid. A report of 2 cases. Acta Cytol. 1993; 37:747-751. Nasuti JF. Utility of the ThinPrep technique in thyroid fine needle aspiration: optimal vs. practical approaches. Acta Cytol. 2006;50:34. Afify AM, Liu J, Al-Khafaji BM. Cytologic artifacts and pitfalls of thyroid fine-needle aspiration using ThinPrep: a comparative retrospective review. Cancer. 2001;93:179-186. Cochand-Priollet B, Prat JJ, Polivka M, et al. Thyroid fine needle aspiration: the morphological features on ThinPrep slide preparations. Eighty cases with histological control. Cytopathology. 2003; 14:343-349. Abati A, Skarulis MC, Shawker T, Solomon D. Ultrasoundguided fine-needle aspiration of parathyroid lesions: a morphological and immunocytochemical approach. Hum Pathol. 1995; 26:338-343. Tseng FY, Hsiao YL, Chang TC. Ultrasound-guided fine needle aspiration cytology of parathyroid lesions. A review of 72 cases. Acta Cytol. 2002;46:1029-1036. Bondeson L, Sandelin K, Grimelius L. Histopathological variables and DNA cytometry in parathyroid carcinoma. Am J Surg Pathol. 1993;17:820-829. Liu F, Gnepp DR, Pisharodi LR. Fine needle aspiration of parathyroid lesions. Acta Cytol. 2004;48:133-136. Abdelghani R, Noureldine S, Abbas A, Moroz K, Kandil E. The diagnostic value of parathyroid hormone washout after fine-needle aspiration of suspicious cervical lesions in patients with hyperparathyroidism. Laryngoscope. 2013;123:1310-1313. Giusti M, Dolcino M, Vera L, et al. Institutional experience of PTH evaluation on fine-needle washing after aspiration biopsy to locate hyperfunctioning parathyroid tissue. J Zhejiang Univ Sci B. 2009;10:323-330. Absher KJ, Truong LD, Khurana KK, Ramzy I. Parathyroid cytology: avoiding diagnostic pitfalls. Head Neck. 2002;24:157-164.

Cancer Cytopathology

September 2014

Cytologic features of parathyroid fine-needle aspiration on ThinPrep preparations.

Previous studies have provided cytologic criteria that aid in the recognition of parathyroid tissue on aspirate smears, including high cellularity, th...
360KB Sizes 1 Downloads 3 Views