PhiliPPine Journal of otolaryngology-head and neck Surgery Vol. 34 no. 2 July – december 2019 PhiliPPine Journal of otolaryngology-head and neck Surgery 2524 PhiliPPine Journal of otolaryngology-head and neck Surgery ORIGINAL ARTICLES ABSTRACT Objective: To determine the association of thyroid malignancy and the following ultrasonographic findings: presence of solid hypoechoic nodule, irregular margins (infiltrative, microlobulated, or speculated), microcalcifications or disrupted rim calcifications with small extrusive hypoechoic soft tissue component, taller than wide shape of the thyroid nodule, and evidence of extrathyroidal extension. Methods: Design: Cross-Sectional Study Setting: Tertiary Government Training Hospital Participants: Records of patients admitted to the Otorhinolaryngology- Head and Neck Surgery ward with a diagnosis of nodular non-toxic goiter, multinodular non-toxic goiter, and thyroid malignancy who underwent thyroid surgery between January 2017 and June 2018 were considered for inclusion. Results: A total of 33 patients, 7 males and 26 females, were included in this study. Patients’ age ranged from 26 to 69 years with an average of 46 years. Thirteen (39.4%) had malignant while 20 (60.6%) had benign histopathologic results. There was a significant association between presence of solid hypoechoic nodule (Fisher exact, n = 33, p = .0047), irregular margins and microcalcifications with malignant histopathology results (X2 (df = 1, N = 33) = 8.57, p = .003). No significant difference was noted in the proportion of subjects with malignancy according to taller than wide nodules (Fisher exact, n = 33, p = 1.000) or presence of extrathyroidal extension nodules or presence of extrathyroidal extension (Fisher exact, n = 33, p = .276). On multivariate analysis using logistic regression, only microcalcification was found to be a significant predictor of malignancy (OR = 8.96, 95% CI: 1.02 – 87.19, p = .05). Conclusion: There was a significant association between presence of solid hypoechoic nodule, margins and microcalcifications with thyroid carcinoma. Only microcalcification was found to be a significant predictor of thyroid malignancy on ultrasound, although our confidence interval was broad. Keywords: calcifications; thyroid; thyroid carcinoma; thyroid nodule; ultrasonography Association of Ultrasonographic Findings and Thyroid Malignancy: A Cross-Sectional Study Clint S. Trinchera, MD Emmanuel Tadeus S. Cruz, MD Quezon City General Hospital Department of Otorhinolaryngology Head and Neck Surgery Correspondence: Dr. Emmanuel Tadeus S. Cruz Department of Otorhinolaryngology Head and Neck Surgery Quezon City General Hospital Seminary Rd., Bgy. Bahay Toro, Quezon City 1106 Philippines Phone: (02) 8863 0800 local 401 Email: orl_hns_qcgh@yahoo.com The authors declared that this is an original manuscript that has not been published or accepted for publication locally or elsewhere, in print or in electronic media. The manuscript has been approved by all the authors, and that the requirements for authorship have been met. Authors practiced honesty and integrity in this work. Disclosures: The authors signed disclosures that there are no financial grants or relationships, intellectual passion, religious, cultural, or political beliefs, and institutional affiliations that may bring about conflict of interest. Presented at the Philippine Society of Otolaryngology-Head and Neck Surgery, Analytical Research Contest (2nd Place). October 23, 2018. Roma Salon, The Manila Hotel. Manila. Presented at the 62nd PSO-HNS Annual Convention & IFHNOS World Tour Manila Analytical Research Contest (2nd Place), December 2018. Marquee tent, EDSA Shangri-La Hotel, Manila. Philipp J Otolaryngol Head Neck Surg 2019; 34 (2): 24-28 c Philippine Society of Otolaryngology – Head and Neck Surgery, Inc. Creative Commons (CC BY-NC-ND 4.0) Attribution - NonCommercial - NoDerivatives 4.0 International PhiliPPine Journal of otolaryngology-head and neck Surgery Vol. 34 no. 2 July – december 2019 PhiliPPine Journal of otolaryngology-head and neck Surgery 2524 PhiliPPine Journal of otolaryngology-head and neck Surgery ORIGINAL ARTICLES Thyroid cancer accounts for approximately 0.5% of all malignancies.1 Though most palpable nodules are benign, there is a 5-10% risk of having a malignancy.2 The American Thyroid Association (ATA) strongly recommends thyroid ultrasonography (UTZ) for patients with suspected or incidentally-detected nodules. This modality is able to determine the presence of suspicious nodules, their character and location, and guide the accuracy of Fine Needle Aspiration Biopsy.3 A number of studies have been published associating specific sonographic findings with thyroid malignancy.1,2,4-9 Many elucidate the role of nodule size and presence of microcalcifications.2,4-9 The ATA listed presence of disrupted rim calcifications with small extrusive hypoechoic soft tissue component, evidence of extrathyroidal extension, nodule shape and features as determinants of malignancy. We initiated this study to determine whether these ultrasonographic features of malignancy are also present in, or if they differ in the local setting. This study aims to determine the association of thyroid malignancy and the following ultrasonographic findings: presence of solid hypoechoic nodule, irregular margins (infiltrative, microlobulated, or speculated), microcalcifications or disrupted rim calcifications with small extrusive hypoechoic soft tissue component, taller than wide shape of the thyroid nodule, and evidence of extrathyroidal extension. METHODS With Institutional Review Board approval, this cross-sectional study was conducted from May to July 2018. Records of patients who were admitted to the Otorhinolaryngology- Head and Neck Surgery ward with a diagnosis of nodular non-toxic goiter, multinodular non-toxic goiter, or thyroid malignancy and underwent thyroid surgery from January 2017 to June 2018 were considered for inclusion. The following were excluded: post-surgery patients previously diagnosed with thyroid malignancy who were managed for tumor recurrence or relapse, and those with no ultrasonographic result or final histopathology results on record. A sample size of 24 was computed using this formula: where: Z α = 95% confidence level= 1.96 Z β = 80% power of study= 0.84 P 1 = estimated proportion of malignant histopathological results among those with malignant ultrasound findings= 30/51 = 58.8 = 0.588 Q1= 1- P 1 = 1- 0.588 = 0.412 P 2 = estimated proportion of malignant histopatholigical results among those with benign ultrasound findings = 15/75= 20.0= 0.20 Q 2 = 1 - P 2 = 1- 0.20 = 0.80 q= 1- p = 1 - 0.394 = 0.606 The clinical histories, physical examination, and histopathologic results recorded in the hospital charts were extracted by the researcher and collated using MS Excel for Mac 2011 version 14.4.2 (Microsoft Corp., Redmond, WA, USA). A senior radiology resident blinded to the final histopathology viewed each ultrasound image and reviewed each printed report to ascertain the presence or absence of the parameters which may not be covered in the official interpretation of the ultrasound results. This was done to standardize the interpretation. If the feature/s were absent on the printed report, the image was reviewed; if both the printed report and the image was absent, then the parameter/s were ruled out. We tabulated the consolidated data using the same software. The presence and type of malignancy were included in the tabulation. Analysis of data Data were encoded and tallied in Statistical Package for Social Sciences (SPSSX 3) version 10 for windows (SPSS Corp., Chicago, IL, USA). Descriptive statistics were generated for all variables. For nominal data frequencies and percentages were computed. For numerical data, mean ± SD were generated. Analysis of the different variables was done using the following test statistics: Chi-square test was used to compare/associate nominal (categorical) data, Fisher Exact test was used when there were expected frequencies <5, Logistic Regression was used in predicting a dichotomous outcome variable (i.e. malignant histopathology result). RESULTS Out of 45 records retrieved, a total of 33 were included in the study. Of the 12 excluded records, 9 had no histopathology result while 8 had no ultrasound report. Of the 33 included, 7 (21.2%) were males and (Zα √ 2pq + Ζβ √ P1Q1 + P2Q2) 2 (P1-P2) 2 n = P1+P2 0.588+0.20 2 2 p= = = 0.394 (1.96 √ 2(0.394)(0.606) + 0.84 √ (0.588 (0.412) + 0.20 (0.80)) 2 (0.588 - 0.20) 2 n = (1.35 + 0.53)2 0.15 n = = 24 PhiliPPine Journal of otolaryngology-head and neck Surgery Vol. 34 no. 2 July – december 2019 PhiliPPine Journal of otolaryngology-head and neck Surgery 2726 PhiliPPine Journal of otolaryngology-head and neck Surgery ORIGINAL ARTICLES 26 (78.8%) were females. The age ranged from 26 to 69 years with an average of 46 ± 13.6 years. The age of patients with malignant results (52.6 ± 15.6 years) was significantly higher (t (df = 31) = 2.42, p = 0.02) than those with benign histopathology (41.7 ± 10.4 years). On the other hand, no significant difference was noted in the proportion of males and females according to histopathology result (Fisher exact, n = 33, p = .58). A solid hypoechoic nodule was present in 21 (63.6%), irregular margins in 9 (27.3%), microcalcifications in 15 (45.5%), taller than wide nodules in 29 (87.9%), and extrathyroidal extension in 4 (12.1%). (Table 2) Of the 33 patients, 13 (39.4%) had thyroid carcinoma while 20 (60.6%) had benign lesions. Malignant histopathologies included 12 with papillary thyroid carcinoma and one with medullary carcinoma while benign lesions comprised 20 colloid adenomatous goiters. There were significant associations noted between histopathology results and the following ultrasonographic findings: 1. presence of solid hypoechoic nodule (Fisher exact, n = 33, p = .047), irregular margins and 2. microcalcifications (X2 (df = 1, N = 33) = 8.57, p = .003). A significantly higher proportion (Fisher exact, n = 33, p = .047) of subjects with solid hypoechoic nodule had malignant histopathology results compared to those with nodules that were not solid with 11 (52.4%) and 2 (16.7%), respectively. Similarly, significantly higher proportions of nodules with irregular margins were malignant compared to those with regular margins with 6 (66.7%) and 7 (29.2%), respectively. Moreover, a significantly higher proportion (X2 (df = 1, N = 33) = 8.57, p = .003) of nodules with microcalcifications yielded malignant results compared to those without microcalcifications with 10 (66.7%) and 3 (16.7%), respectively. On the other hand, no significant difference was noted in the proportion of subjects with malignancy according to taller than wide nodules (Fisher exact, n = 33, p = 1.000) or presence of extrathyroidal extension (Fisher exact, n = 33, p = .276). On multivariate analysis using logistic regression, only microcalcification was found to be a significant predictor of malignancy. The risk of patients with microcalcification for thyroid malignancy was almost 9 times higher than those without microcalcification (OR = 8.96; 95% CI = 1.02 – 87.19; p = .05). (Table 4) DISCUSSION This study revealed a significant association between malignant histopathologic results of the thyroid and the presence of solid hypoechoic nodules, irregular margins and microcalcifications on ultrasound. Conversely, no significant association was found between thyroid malignancy and taller than wide nodules or presence of extrathyroidal extension on ultrasound. Table 2. Distribution According to Ultrasonographic Findings Frequency (n=33) Percentage Presence of solid hypoehoic nodule Solid Not solid Margins Irregular Regular Microcalcifications With Without Taller than wide nodules Taller than wide Not taller than wide Extrathyroidal Extension With extension Without extension 21 12 9 24 15 18 29 4 4 29 63.6 36.4 27.3 72.7 45.5 54.5 87.9 12.1 12.1 87.9 Table 1. Comparison of Demographic Characteristics According to Histopathology Results Histopath Result Malignant Benign (n=13) (n=20) Total p-value* Age (in years) Mean ± SD Sex Male Female 52.62 ± 15.61 3 (42.9%) 10 (38.5%) 41.70 ± 10.41 4 (57.1%) 16 (61.5%) 46.00 ± 13.60 7 26 .02 (S)† 1.00 (NS)‡ * p>0.05- Not significant; p ≤0.05-Significant Data presented as Mean ± SD, (medians) were computed as needed; or as frequency (%) † T-test; ‡Fisher Exact test * S- Significant; (NS)- Not Significant These findings are similar to those of Anil and colleagues wherein solid nodules on UTZ had the highest sensitivity (69-75%) for malignancy in comparison to cystic or spongiform appearance.7 Cystic nodules were more likely benign (with only about 2% malignant),8 similar to our findings. Solid nodules are a common ultrasound feature of papillary thyroid carcinoma in around 70% of cases,4 anaplastic and medullary thyroid carcinoma also display this feature.4 Irregular margins may be infiltrative, microlobulated or spiculated.3 Irregularity of margins is attributed to the nodule losing its pseudocapsule of fibrous connective tissues making the tumor and PhiliPPine Journal of otolaryngology-head and neck Surgery Vol. 34 no. 2 July – december 2019 PhiliPPine Journal of otolaryngology-head and neck Surgery 2726 PhiliPPine Journal of otolaryngology-head and neck Surgery ORIGINAL ARTICLES malignancy (>90%). In a study by Ren and colleagues (2015), taller than wide shape is a good predictor of papillary thyroid carcinoma in small nodules with high sensitivity and specificity (81.4% and 96.8%, respectively).9 However, the sensitivity was significantly low and odds ratios were smaller (1.6-2.9) when assessed separately from other features.3 In contrast, in this study, no significant association was observed between taller than wide shape and malignancy. The size of the nodule was excluded because this was not included in the parameters set by ATA. No qualifications were made as to large (>1cm) or small (<1); only the anteroposterior to transverse ratio was considered. Extension to adjacent structures and metastasis to lymph nodes are specific signs of thyroid cancer.7 Papillary thyroid cancer is known to metastasize to the pre/paratracheal group of lymph nodes.4,7 Anaplastic carcinoma displays extracapsular and vascular spread (33%). Medullary carcinoma may present with lymph node and distant organ metastasis.7 In this study, 3 had extrathyroidal invasion; 2 with extension to lateral cervical lymph nodes, and 1 to the submandibular region. However, no significant association was found between extrathyroidal invasion and malignancy. This may be due to the small number of subjects with extrathyroidal invasion which may differ in a larger population since sample size is inversely proportional to p value and CI. In this study, only microcalcification proved to be a significant predictor of malignancy using multivariate analysis. The risk of those with microcalcifications for having malignancy was almost 9x higher than those without microcalcification. Microcalcifications were noted to have strong association with malignancy.1,2 In a study by Sio et al., 86% of papillary thyroid carcinoma cases displayed microcalcification on thyroid UTZ.2 This feature had the highest accuracy (76%), specificity (44-95%), and predictive value (77.9%) in detecting malignancy.1 These sonographic findings are consistent with hyperechoic foci with or without shadowing, corresponding histopathologically to psammoma body clusters.1,7 The American Thyroid Association recommends ultrasonography in patients with known or suspected thyroid nodules. Asymptomatic nodules can be found in as many as 50% of adults1,6 while thyroid malignancy affects 1/10,000 people annually.1 The challenge lies in identifying whether a nodule is benign or malignant, and who among the population would necessitate early aggressive management or avoidance of unnecessary investigation.4 At present, the ATA categorized nodules to be of high suspicion (malignancy risk >70% to 90%) if the following are present in thyroid ultrasonography: solid hypoechoic nodule, irregular margins (specified as either infiltrative, microlobulated, or speculated), microcalcifications or disrupted rim Table 3. Distribution According to Ultrasonographic Findings and Histopath Results Histopath Result Malignant Benign (n=13) (n=20) Total p-value* Presence of solid hypoehoic nodule Solid Not solid Margins Irregular Regular Microcalcifications With Without Taller than wide nodules Taller than wide Not taller than wide Extrathyroidal Extension With extension Without extension 11 (52.4%) 2 (16.7%) 6 (66.7%) 7 (29.2%) 10 (66.7%) 3 (16.7%) 11 (37.9%) 2 (50.0%) 3 (75.0%) 10 (34.5%) 10 (47.6%) 10 (83.3%) 3 (33.3%) 17 (70.8%) 5 (33.3%) 15 (83.3%) 18 (62.1%) 2 (50.0%) 1 (25.0%) 19 (65.5%) 21 12 9 24 15 18 29 4 4 29 .05 (S)‡ .05 (S)‡ .003 (S)§ 1.00 (NS)‡ .28 (NS)‡ * p>0.05- Not significant; p ≤0.05-Significant Data presented as Mean ± SD, (medians) were computed as needed; or as frequency (%) ‡ Fisher Exact test; §Chi-square test * S- Significant; (NS)- Not Significant Table 4. Multivariate Analysis of the Different Ultrasonographic Findings Predictive of Malignant Histopath Results Variable OR 95% CI P value Age Sex (male) Solid nodule Irregular margins Microcalcifications Taller than wide nodules Extrathyroidal extension 1.08 0.53 2.03 1.83 8.96 0.07 9.50 0.99 – 1.18 0.03 – 9.79 0.14 – 28.05 0.17 – 19.79 1.02 – 87.19 0.002 – 2.91 0.49 – 182.82 .07 (NS) .67 (NS) .60 (NS) .62 (NS) .05 (S) .16 (NS) .14 (NS) Logistic Regression Analysis the thyroid parenchyma indistinct.7 Results of this study showed that significantly higher proportions of nodules with irregular margins were malignant compared to those with regular margins. This feature may also be found in benign nodules with a low sensitivity (48.3- 55%) but with high specificity (83-91%) and predictive value (60-81.3%).7 The ATA guidelines also implicated taller than wide thyroid nodules measured in transverse view having high specificity for detecting PhiliPPine Journal of otolaryngology-head and neck Surgery Vol. 34 no. 2 July – december 2019 PhiliPPine Journal of otolaryngology-head and neck Surgery 2928 PhiliPPine Journal of otolaryngology-head and neck Surgery ORIGINAL ARTICLES REFERENCES 1. Smith-Bindman R, Lebda P, Feldstein VA, Sellami D, Goldstein RB, Brasic N, et al. Risk of thyroid cancer based on thyroid ultrasound imaging characteristics: results of a population-based study. JAMA Intern Med. 2013 Oct; 173(19): 1788-1796. DOI: 10.1001/jamainternmed.2013.9245; PMID: 23978950 PMCID: PMC3936789. 2. Sio MCD, Uy JAU, Soriano RG. Calcifications in Thyroid Ultrasonography and Thyroid Carcinoma. Philipp J Otorhinolaryngol Head Neck Surg. 2014; 29(2): 15-18. 3. Haugen BR, Alexander EK, Bible KC, Doherty GM, Mandel SJ, Nikiforov YE, et al. 2015 American Thyroid Association Management Guidelines for Adult Patients with Thyroid Nodules and Differentiated Thyroid Cancer. The American Thyroid Association Guidelines Task Force on Thyroid Nodules ad Differentiated Thyroid Cancer. Thyroid. 2016 Jan; 26(1): 1-133. DOI: 10.1089/ thy.2015.0020; PMID: 26462967 PMCID: PMC4739132. 4. Wong KT, Ahuja AT. Ultrasound of thyroid cancer. Cancer Imaging. 2005 Dec 9; 5: 157-166. DOI: 10.1102/1470-7330.2005.0110; PMID: 16361145 PMCID: PMC1665239. 5. Hershman JM. Malignancy of a thyroid nodule can be predicted by ultrasonography if it has microcalcifications and is solid and larger than 2cm. Clin Thyroidol. 2013 Nov; 25(11): 256-258. 6. Vera MI, Merono T, Urrutia MA, Parisi C, Morosan Y, Rosmarin M, et al. Differential profile of ultrasound findings associated with malignancy in mixed and solid thyroid nodules in an elderly female population. J Thyroid Res. 2014; 2014: 761653. DOI: 10.1155/2014/761653; PMID: 25050189 PMCID: PMC4094854. 7. Anil G, Hegde A, Chong FH. Thyroid nodules: risk stratification for malignancy with ultrasound and guided biopsy. Cancer Imaging. 2011 Dec 28; 11: 209-223. DOI: 10.1102/1470- 7330.2011.0030; PMID: 22203727 PMCID: PMC3266587. 8. Hoang J. Thyroid nodules and evaluation of thyroid cancer risk. Australas J Ultrasound Med. 2010 Nov. 13(4): 33-36. DOI: 10.1002/j.2205-0140.2010.tb00177.x; PMID: 28191095 PMCID: PMC5024877. 9. Ren J, Liu B, Zhang LL, Li Hy, Zhang F, Li S, et al. A taller-than-wide shape is a good predictor of papillary thyroid carcinoma in small solid nodules. J Ultraosund Med. 2015 Jan; 34(1):19-26. DOI: 10.7863/ultra.34.1.19; PMID: 25542935. calcifications with small extrusive hypoechoic soft tissue component, taller than wide shape of the thyroid nodule and evidence of extrathyroidal extension. Our study has several limitations. Although the required number of subjects of 24 was met, the study sample may not fairly represent the larger population because of our convenience sampling technique. Random or consecutive sampling with clearer inclusion and exclusion criteria in future studies may address this. Because this is an observational non-randomized study, our benign and malignant “groups” may also vary in prognosis due to different demographics, illness severity or other baseline characteristics. Moreoever, we have at least two potential sources of measurement or observer bias – the histopathologic readings (taken at face value from available reports), and the ultrasonographic readings (despite verification by a blinded senior radiology resident). Future studies can improve on this with random and concealed allocation of participants to groups and prospectively-obtained masked outcome assessment using objective and validated measures by independent, blinded observers. In conclusion, our study found a significant association between malignant histopathologic results of the thyroid and presence of solid hypoechoic nodule, margins and microcalcifications but noted no significant difference with taller than wide nodules or presence of extrathyroidal extension on ultrasound. Only microcalcification was found to be a significant predictor of malignancy using multivariate analysis. The risk of patients with microcalcification for malignant result was almost 9 times higher than those without microcalcification, but this requires careful scrutiny, as our confidence interval was broad.