Secondary Logo

Journal Logo

Original Article

Value of Combined Cytomorphological Parameters in Improving Diagnostic Accuracy of Papillary Thyroid Carcinoma on Cytology - The Five Dependable Features

Kumari, Neha; Kalonia, Tushar; Malik, Akanksha,; Kumar, Arvind; Rao, Shalinee

Author Information
Indian Journal of Endocrinology and Metabolism: Sep–Oct 2021 - Volume 25 - Issue 5 - p 402-409
doi: 10.4103/ijem.ijem_389_21
  • Open

Abstract

INTRODUCTION

Thyroid carcinoma is the most common malignancy of endocrine origin.[1] Papillary thyroid carcinoma (PTC) represents the most common thyroid gland malignancy accounting for 80%–90% of all thyroid cancers.[2] In the evaluation of thyroid swellings, fine-needle cytology (FNC)/fine-needle aspiration cytology (FNAC) forms the preliminary test for diagnosis. Owing to its fairly high sensitivity, specificity, cost-effectiveness, and simplicity, FNC/FNAC is a promising and efficient means of distinguishing patients who require therapeutic surgical intervention from those who may require just long-term follow-up.

There are precise cytological features of PTC as recognized by Synopsis of National cancer institute (NCI), Thyroid Fine-Needle Aspiration State of the Science Conference.[3] However, few cases still pose problems during cytological evaluation, and in up to one-quarter of the cases, they are misclassified as lesions other than PTC.[4] These cases with overlapping features or lacking clear-cut morphology are difficult to recognize as PTC. It results in bothersome rates of false-positive and false-negative cytology reports with serious implications in patient care and management. False-positive cytology report for PTC is due to shared cytological features with other benign and malignant entities in the thyroid. Absence or infrequent presence of these features, scant cellularity, and degenerative changes can result in a false-negative diagnosis, at times, render diagnoses like possibility of PTC/rule out PTC/suspicious of PTC, which often creates a diagnostic dilemma in the clinician's mind regarding the management protocol for these cases.[5] Thus, this study aimed to determine the reliability of different cytological features individually and in combinations to identify any individual/set of cytological features carrying the maximum ability to correctly diagnose PTC and distinguish it from its benign mimickers as both have very different management protocols.

MATERIALS AND METHODS

After obtaining approval from the institute's ethics committee, an observational study was carried out in a tertiary care institute in Rishikesh, Uttarakhand. Cytology and histopathology slides of patients who underwent surgical thyroid resection (hemithyroidectomy/total thyroidectomy) between January 2018 and June 2020 with preoperative FNAC were screened.

Inclusion criteria

All histopathologically proven cases of PTC were taken out and their FNAC slides were reviewed for assessing cytology features.

Exclusion criteria

Histopathologically proven cases of PTC for which FNAC slides were not available or were of poor staining quality or contained inadequate cellularity were excluded from the study.

All cases included in our study were divided into the following two groups.

Group 1: This study group included 36 cases whose diagnoses were confirmed as PTC on histopathology. In cytology, these cases were diagnosed as PTC (20 cases), suspicious of PTC (5 cases), possibility of PTC cannot be ruled out (1 case), follicular neoplasm (2 cases), atypia of undetermined significance (AUS) (2 cases), adenomatous goiter (5 cases), and lymphocytic thyroiditis (1 case).

Group 2: This control group included 38 cases with postoperative confirmed benign histopathological diagnosis of adenomatous goiter/hyperplastic nodule/lymphocytic thyroiditis, and in cytology, they were diagnosed as adenomatous goiter (25 cases), lymphocytic thyroiditis (6 cases), follicular neoplasm (4 cases), possibility of PTC (1 case), and AUS (2 cases).

Stained cytology slides of all cases of groups 1 and 2 were taken out from the archive and assessed by three pathologists together (NK, TK, AM). Four slides, two each of May–Grunwald Giemsa (MGG) stained and two of Papanicolaou (Pap) stained slides were reviewed as per Bethesda system of reporting and assessed for 12 cytological features of PTC in each case.[5] The two groups were then compared with respect to demographic profile and cytology features.

Parameters taken into consideration included all the major as well as minor diagnostic cytological features of PTC as established by NCI FNA Science Conference State.[3] Cytological features assessed included the following:

  • (i) Papillae with central fibrovascular cores, 3-D caps, nuclear grooves (NG), intranuclear cytoplasmic inclusions (INCI), giant cells, cystic macrophages, cellular swirls, psammoma bodies, pale chromatin, nuclear overlapping, nuclear enlargement/irregularity, and metaplastic cells.
  • (ii) NG and INCI were semi-quantitatively assessed based on their frequency in the smears and graded as Grade I- many/frequent throughout the smear, Grade II- frequent in occasional groups of cells, and Grade III- occasional throughout the smear/absent.
  • (iii) Nuclear enlargement was described as being absent or present based on a subjective assessment of the nuclear size. All other features were also graded as present or absent.

Statistical analysis

Categorical variables were presented in number and percentage (%), and continuous variables were presented as mean ± SD and median. Normality of data was tested using Kolmogorov–Smirnov test. If the normality was rejected, then a nonparametric test was used.

Statistical tests were applied as follows:

  1. Quantitative variables were compared using Independent t test/Mann–Whitney Test (when the data sets were not normally distributed) between the two groups.
  2. Qualitative variables were compared using Chi-Square test/Fisher's Exact test.
  3. Receiver operating characteristic curve was used to find out the cut-off point of several cytological features to predict PTC.
  4. Sensitivity, specificity, PPV, and NPV of cytological features were calculated to predict PTC.

P < 0.05 was considered statistically significant.

The data was entered in an MS Excel spreadsheet and analysis was done using Statistical Package for Social Sciences (SPSS) version 21.0.

RESULTS

This study included 74 cases distributed in group 1 or group 2 with a median age of 34 and 35 years, respectively [Table 1]. Overall, there were 60 females and 14 males. Significant difference was observed between the two patient groups for gender ratio but not for age [Table 1].

T1
Table 1:
Comparison of demographic characteristics between groups 1 and 2

Distribution of cases as per Bethesda categories is depicted in [Table 2] and [Figure 1]. Comparison of various cytological features between study group 1 and control group 2 is featured in Table 3. In group 1, 55.5% (20/36) of the cases were diagnosed as PTC on cytology, 16.6% (6/36) were labelled as suspicious/suggestive of PTC, whereas in 27.7% (10/36) cases the possibility of PTC was not suggested. Of these 27.7% cases, 11.11% cases were diagnosed as either follicular neoplasm or AUS and 16.6% were given a benign diagnosis [Table 3].

T2
Table 2:
Categorization of groups 1 and 2 based on Bethesda system of thyroid reporting
F1
Figure 1:
Comparison between groups 1 and 2 based on Bethesda category
T3
Table 3:
Comparison of cytologic features between groups 1 and 2

The cytological features that showed statistically significant differences between group 1 and group 2 were papillae with cores, 3D-caps, cellular swirls, nature of colloid, NG, INCI, squamous metaplastic cells, and cytoplasmic vacuolations [Table 3]. Cystic/foamy macrophages, nuclear overlapping and psammoma bodies did not differ significantly between groups 1 and 2 [Table 3]. Receiver operating characteristic curve analysis to predict PTC revealed that the presence of ≥5 cytological features, namely papillae with cores, cellular swirls, NG, INCI, and psammoma bodies, when found together in a case could diagnose PTC on cytology correctly (PPV) in 78.95% of the cases, with a negative predictive value of 83.33% [Table 4 and Figure 2]. Diagnostic accuracy of these five features combined was 81.08% [Table 5]. Papillae with cores and nuclear grooving were the most sensitive cytological features (sensitivity: 61.1%) whereas INCI followed by cellular swirls and NG (100%, 97.3%, and 97.3%, respectively) were the most specific features to diagnose PTC [Table 5]. Psammoma bodies had high specificity but very low sensitivity to detect PTC on cytology. Table 5 shows the sensitivity, specificity, PPV, and NPV of five cytological features to predict PTC.

T4
Table 4:
Receiver operating characteristic curve with a combination of five cytological features to predict PTC
F2
Figure 2:
Receiver operating characteristic curve on combining five cytological features to predict PTC
T5
Table 5:
Sensitivity, specificity, PPV, and NPV of cytological features to predict PTC

Multinucleated giant cells (MGC) in benign condition were small ovoids having pale/foamy cytoplasm and few nuclei [Figure 3]. In contrast, MGC found in PTC cases were of two types: MGC as found in benign and MGC that appeared larger with varying shape, more nuclei, and dense cytoplasm. Clinical and histopathological parameters of group 1 are described in Table 6.

F3
Figure 3:
Comparison of the number of nuclei in giant cells between groups 1 and 2 (nonparametric variable, Box whisker plot)
T6
Table 6:
Clinical and histological details of PTC cases included in the study

DISCUSSION

Worldwide, thyroid cancer ranks ninth in incidence and is three times more common in women than in men.[6] It is the most common type of thyroid cancer accounting for 70%–80% of all thyroid malignancies and compared to other thyroid cancers carries the best prognosis.[7] Since 1980, incidence of thyroid carcinoma has shown an increase in trend throughout the world, which may be attributed to increased detection of PTC through improved diagnosis and changing prevalence of risk factors or possibly may be an impact of overdiagnosis.[89] However, as over the same duration, death rate due to PTC has not shown any significant change, one pertinent question which comes to mind is whether the rise in PTC cases is actually true or it is an overdiagnosis.

Cytological evaluation is a well-established diagnostic modality and a variable number of cytological features have been used by different studies to diagnose PTC on FNAC. In the study conducted by Chandanwale et al.,[10] frequency of papillae, NG, and INCI were significantly higher in PTC cases compared to those that were diagnosed as suggestive of PTC. Wu et al.[11] reported nuclear enlargement, fine chromatin, NG, INCI, flat syncytial sheets, and some amount of thick colloid as the most common cytological features of PTC. Castro-Gomez et al.[12] reported the presence of tridimensional fragments, anisonucleosis, nuclear bars, powdery chromatin, cytoplasmic vacuoles, metaplastic cytoplasm, and autolysis as cytological features that occurred significantly higher in the PTC group as compared to the benign thyroid lesions. Castro-Gomez et al.[12] noted that papillae, psammoma bodies, multinucleated giant cells, spindle cells, colloid, and macrophages did not show any statistically significant between the benign and PTC groups.

Kumar et al.[13] reported powdery chromatin, nuclear inclusions, nuclear grooves, and cellular swirls as the most consistent features present in all their cases of PTC, with nuclear grooves being more abundant than nuclear inclusion; 14% of their cases showed frequent nuclear grooves (>60%/10HPF), and 64% of the cases showed 2+ nuclear grooves (i.e., 31%–60%/10HPF), whereas INCI were infrequently identified (0%–30%/10HPF) in 93% of their cases. The majority of our cases (61%) showed frequent nuclear grooves throughout the smear while 30% showed frequent INCI [Figure 4].

F4
Figure 4:
Microphotograph showing intracytoplasmic inclusion (arrow) (May Grunwald Giemsa, ×400)

Das et al.[14] observed that the presence of ≥3 of the following features—psamomma bodies, papillae, inclusions, nuclear grooves, and granular chromatin—facilitate cytological diagnosis of papillary carcinoma of thyroid, with frequent grooves and inclusions being the most dependable. Miller et al.[15] studied logistic regression analysis of the cytological variables and found that a combination of intranuclear cytoplasmic pseudo inclusion, papillary structure without adherent blood vessels, and dense metaplastic cytoplasm were the most important variables with 100% predictive value when a combination of any of these two was found. In the current study, by using the logistic regression analysis, we found that combination of five cytological features, namely papillae with cores, cellular swirls, NG, INCI, and psammoma bodies, had a diagnostic accuracy of 81% and could correctly diagnose (PPV) PTC in 78.9% of cases [Table 7]. PPV for cellular swirls and INCI was 100%, and it exceeded 95% when cellular swirls, NG, and INCI were present together.

T7
Table 7:
The Pentad of Cytological Features

Cytological features of PTC are not specific as they are also seen in other malignant and benign lesions of the thyroid. Nuclear grooves can be seen in medullary thyroid cancer, follicular cancer, follicular adenoma, lymphocytic thyroiditis, and nodular goiter.[1617] Certain studies have shown that quantification of intranuclear grooves can be helpful as PTC tends to have a greater number of cells with intranuclear grooves in comparison to other thyroid lesions and widespread pattern of NG can be a useful criterion.[1819] It has also been noted that in benign entities, usually, nuclear grooves are thin and/or incomplete, whereas they are thick longitudinal and accompanied by other cytological features in cases of PTC.[18] INCI can be seen in medullary carcinoma, insular carcinoma, hyalinizing trabecular adenoma, Hashimoto's thyroiditis, and multinodular goiter.[1619] In the present study, 10 cases of PTC that were incorrectly diagnosed on cytology as adenomatous goiter, AUS, or follicular neoplasm showed nuclear grooves in nine cases, while INCI was not seen in any of these cases. On further scrutiny of cytosmears, three of the nine cases showed frequent NG (Grade 1), a feature that could have helped in favoring the diagnosis of PTC in these cases. Seven cases showed infrequent/paucity of NG and all 10 cases showed absence of INCI, thereby posing diagnostic difficulty.

Similar observations were made by Renshaw et al.[20] who compared the cytological features of PTC between two groups, group that performed well versus one that performed poorly in the College of American Pathologists (CAP) cytology program. They reported that the group that performed poorly or were difficult to recognize as PTC were significantly more likely to lack nuclear enlargement, intranuclear inclusions, and pale chromatin, whereas nuclear crowding, nuclear overlapping, colloid, cellularity, Hurthle cell changes, and type of staining did not differ significantly between their two study groups.

Another less commonly reported cytological feature as a diagnostic clue for PTC are cellular swirls [Figure 5]. Cellular swirls are concentrically arranged follicular cells with peripheral neoplastic cells whose axis is perpendicular to the radius of the swirl.[21] Cellular swirls have been reported as highly specific for PTC when found in cytological smears.[132223] We also found similar results as the presence of cellular swirls was limited to cases of PTC and none of the cases in group 2 showed cellular swirling. Cellular swirls may be confused with the spherules or whorl-like structures structure, which consists of cells with prominent nuclear changes such as enlarged nuclei and thickened nuclear membranes, and is seen in cases of the cribriform-morular variant of PTC.[24] In contrast, nuclear features in cellular swirls are bland.

F5
Figure 5:
Microphotograph showing cellular swirls (arrow) (May Grunwald Giemsa, ×400)

One case in group 2 was falsely diagnosed as PTC on cytology and turned out to be nodular goiter with lymphocytic thyroiditis on histopathology showing few nuclear grooves (grade 2) and rare INCI (grade 3).

Results of our study and revelations by various studies highlight the fact that cytological features in PTC may vary in some cases, causing diagnostic dilemmas in them owing to either limited morphology or overlapping cytological findings. False-positive and false-negative cytology results in overdiagnosis and underdiagnosis and may misguide the clinician. Thus, it is essential that such focused studies be carried out on a larger scale, and a systematic review and meta-analysis may be helpful in defining precise criteria to identify PTC on cytosmears.

CONCLUSION

The majority of the cytological features of PTC are neither specific nor constant. Incorrect diagnoses on FNAC in suspected cases of papillary carcinoma of thyroid can be minimized if the pathologist is aware of these pitfalls. Relying on a combination of most sensitive features like nuclear grooves and papillae and specific features such as cellular swirls and intranuclear cytoplasmic inclusions rather than on any one single cytological feature alone can help in reducing the false-positive/false-negative diagnoses in papillary thyroid carcinoma.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

REFERRENCES

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
Keywords:

Cellular swirls; papillae; papillary thyroid carcinoma; PTC

Copyright: © 2022 Indian Journal of Endocrinology and Metabolism