Abstract
Objectives
For small thyroid cancers (≤2 cm), tumour volume may better predict aggressive disease, defined by lymphovascular invasion (LVI) than a traditional single measurement of diameter. We aimed to investigate the relationship between tumour diameter, volume and associated LVI.
Methods
Differentiated thyroid cancers (DTC) ≤ 2 cm surgically resected between 2007 and 2016 were analysed. Volume was calculated using the formula for an ellipsoid shape from pathological dimensions. A ‘larger volume’ cut-off was established by receiver operating characteristic (ROC) analysis using the presence of lateral cervical lymph node metastasis (N1b). Logistic regression was performed to compare the ‘larger volume’ cut-off to traditional measurements of diameter in the prediction.
Results
During the study period, 2405 DTCs were surgically treated and 523 met the inclusion criteria. The variance of tumour volume relative to diameter increased exponentially with increasing tumour size; the interquartile ranges for the volumes of 10, 15 and 20 mm diameter tumours were 126, 491 and 1225 mm3, respectively. ROC analysis using volume to predict N1b disease established an optimal volume cut-off of 350 mm3 (area under curve = 0.59, P = 0.02) as ‘larger volume’. ’Larger volume’ DTC was an independent predictor for LVI in multivariate analysis (odds ratio (OR) = 1.7, P = 0.02), whereas tumour diameter > 1 cm was not (OR = 1.5, P = 0.13). Both the volume > 350 mm3 and dimension > 1 cm were associated with greater than five lymph node metastasis and extrathyroidal extension.
Conclusion
In this study for small DTCs ≤ 2 cm, the volume of >350 mm3 was a better predictor of LVI than greatest dimension > 1 cm.
Introduction
Well-differentiated thyroid cancer (DTC) and papillary thyroid cancer (PTC) incidence has been increasing in comparison to other thyroid cancers (Davies & Welch 2014, Davies & Hoang 2021). However, most thyroid cancer diagnoses are unlikely to be life-threatening with a greater than 95% 10-year overall survival (Ito et al. 2018, Park et al. 2018). Most of the rise in new cases seem to be small and lower risk cancers (Guth et al. 2009, Dal Maso et al. 2018). Management of these cancers is controversial, with treatment recommendations ranging from active surveillance (AS) to surgery and the use of radioactive iodine therapy (Boucai et al. 2017, Zanocco et al. 2019). While for some patients more aggressive treatment can be indicated or preferred, it has been argued that with conservative approaches the same or better overall outcomes may be achieved with careful patient selection (Tuttle 2018).
Pre-operative assessment and prediction of high-risk cancers is important when counselling patients to facilitate shared decision-making in deciding on a treatment approach. Currently, the major risk stratification and staging systems used are the 2015 American Thyroid Association (ATA) risk scoring system (Haugen et al. 2016) and the 8th American Joint Committee on Cancer (AJCC) staging systems (Edge 2017), and both use the greatest diameter of the tumour to quantify size. However, DTCs have been reported to be ellipsoid in shape rather than spherical, so tumour volume may be a better representation of a tumour’s size (Pennington et al. 2018). Volume may be a better predictor of the tumour’s pathological characteristics and hence may allow better pre-operative risk stratification. Currently, the ATA guidelines (Haugen et al. 2016) combine clinical and pathological features to estimate recurrence risk.
Pathological findings such as the presence of lymphovascular invasion (LVI) and lymph node metastases often upstage the patient to an intermediate risk with the presence of LVI thought to increase the risk of recurrence by 15–30% (Haugen et al. 2016); hence, completion thyroidectomy is often necessary after review of the initial post-operative histology. While the presence of nodal metastases is a factor that may be able to be determined pre-operatively, LVI is not. So if we are able to better predict for LVI, it may allow for better risk assessment from the time of diagnosis.
Prior research examining DTCs of all sizes found that volume correlates strongly with nodal metastasis, better than the greatest diameter alone (Pennington et al. 2018). Similarly, in AS studies of DTC, it has been established that serial volume measurements can allow the prediction of growth characteristics earlier than measurements of the greatest diameter (Tuttle et al. 2017). Tumour volume studies have also been shown to be an independent predictor of outcomes in other cancers such as early stage rectal cancer (Jiang et al. 2018), non-small cell lung cancer (Takenaka et al. 2016, Su et al. 2017) and prostate cancer (Stamey et al. 1999).
This study aimed to evaluate the relationship of pathological tumour volume to greatest dimension for small (≤2 cm) DTC and assess if volume correlates with post-operative risk factors for recurrence such as LVI.
Methodology
Study design
The study aimed to evaluate whether risk stratification of small DTCs was possible based on tumour volume as opposed to greatest diameter. A retrospective study analysis of consecutive, surgically treated DTC patients at Royal North Shore Hospital was performed after approval was obtained from the local human research ethics committee (Northern Sydney Local Health District). Data were extracted from the prospectively maintained endocrine surgery database and patients were included if they underwent surgery and had pathologically confirmed DTC with a diameter of ≤20 mm. For patients with multifocal cancers, the largest tumour by greatest dimension was used. Patients were excluded from the analysis if they did not have all three dimensions of their tumour histologically reported or were classified as high risk as per the 2015 ATA risk of recurrence guidelines (Haugen et al. 2016).
Operative treatment
All patients included in this study underwent surgery within a tertiary unit with a pathological assessment performed by pathologists with endocrine cancer expertise. Patients also had both known and incidental thyroid cancers, with select patients undergoing central lymph node dissection for either prophylactic or therapeutic intent and if indicated lateral lymph node dissection for therapeutic intent.
Calculating tumour volume
All pathological assessment of tumours was reported independently by our pathologists. Tumour volume was calculated assuming an ellipsoid shape using the diameter method (volume = length (mm) × width (mm) × height (mm) × π/6) as that is the clinical impression of the shape and method used in previous endocrine surgery studies (Tuttle et al. 2017, Pennington et al. 2018).
Outcome measures
The primary endpoint was the presence of LVI, While other risk factors analysed included pathologically confirmed presence of more than five metastatic lymph nodes, presence of lateral cervical nodal metastasis (N1b) and minor extrathyroidal extension (ETE). Minor ETE was defined as spreading that extends focally out of the thyroid into the adjacent soft tissue but not widespread into the soft tissue or skeletal muscle.
Establishing the optimal cut-off
To form a hypothetical volume cut-off, a receiver operating characteristic (ROC) curve analysis was performed for different outcomes including N1b. The optimal cut-off point was determined to be the point closest to the top left corner indicating the optimal combination of sensitivity and specificity, as described in other studies (Unal 2017). A separate subgroup excluding patients with N1b disease was conducted using the initial cut-off to see if volume had the same utility.
Statistical analysis
Statistical tests were conducted using IBM SPSS Statistics Version 27.0 (IBM). To evaluate the associations between clinicopathological characteristics and tumour size, Fisher’s exact test or chi-squared test was used for categorical variables, while the association with age (the only continuous variable) was assessed with the Student’s t-test. Binary logistic regression was conducted at selected endpoints. A histogram was created to illustrate the tumour volume distribution of the cohort. Both linear and exponential regression analyses were performed to explore the relationship between tumour volume and greatest dimension. For all statistical analyses, a two-sided P-value < 0.05 was considered to be statistically significant.
Results
Patient demographics
Between March 2007 and December 2016, a total of 2405 patients underwent surgery for DTC. Patients who had tumours with the greatest dimension > 20 mm (n = 708), those who did not have all three dimensions pathologically reported (n = 1164), or those who were classified as high-risk as per the 2015 ATA risk of recurrence (n = 10) were excluded. A total of 523 patients were identified for final analysis (Supplementary Fig. 1, see section on supplementary materials given at the end of this article) with included patients’ demographics summarised in Table 1.
Baseline patient demographics and clinicopathological characteristics.
Characteristics | Number of patients |
---|---|
Age (years) mean age 50.0, range 13–85 |
|
≤55 | 195 (37.3%) |
>55 | 328 (62.7%) |
Sex | |
Male | 101 (19.3%) |
Female | 422 (80.7%) |
Hashimoto’s thyroiditis | |
Yes | 38 (7.3%) |
No | 485 (92.7%) |
Type of operation | |
Thyroidectomy | 500 (95.6%) |
Hemithyroidectomy | 23 (4.4%) |
Histological type | |
Papillary cancer | 488 (93.3%) |
Follicular cancer | 24 (4.6%) |
Hurthle cell cancer | 11 (2.1%) |
Incidental | |
Yes | 431 (82.4%) |
No | 92 (17.6%) |
Lymphovascular invasion | |
No | 385 (73.6%) |
Yes | 138 (26.4%) |
Total positive lymph nodes | |
≤5 | 464 (88.7%) |
>5 | 59 (11.3%) |
Lateral lymph node disease (N1b) | |
No | 454 (86.8%) |
Yes | 69 (13.2%) |
Minor extrathyroidal extension | |
No | 418 (79.9%) |
Yes | 105 (20.1%) |
The mean age of the cohort was 50 years (range: 13–85), with 422 (80.7%) patients being female. Total thyroidectomy was performed on 500 patients (96%) while the remaining 23 (4%) had a hemithyroidectomy. A total of 397 (75.9%) patients had central lymph node dissection and 181 (34.6%) had lateral lymph node dissection. There were 431 patients (82.4%) who had pre-operatively known DTC and the remaining 92 (17.6%) had incidental DTC found after operation for other indications (Table 1). The major histological type diagnosed was papillary cancer (93%) followed by follicular thyroid cancer (5%) and Hurthle cell carcinoma (2%).
A total of 170 patients (32.5%) had a micro-DTC (≤10 mm) and 353 (67.5%) with greatest dimension between 10 and 20 mm. Both patients treated for cancers (≤10 mm) and 10–20 mm were similar except for a greater number of patients treated with total thyroidectomy with larger cancers (Supplementary Table 1).
Relationship of greatest diameter to tumour volume
Tumour volume ranged from 0.3 to 3780.4 mm3, with a median of 494.8 mm3 (with Q1–Q3 209.4–1047.2 mm3). The distribution of volumes was positively skewed as shown in the histogram plot (Fig. 1). Regression analysis between tumour volume and greatest dimension, and the correlation analysis of both exponential and linear model fits are presented in Fig. 2. The results suggest that the correlation between tumour volume and greatest dimension is better explained using an exponential growth model compared to linear model (r2 = 0.771 vs 0.624).
As the greatest dimension increases, the variance between tumour volume and tumour diameter increased, with inter-quartile ranges (IQR) for volume measurements of 126, 491 and 1225 mm3 for tumour diameters of 10, 15 and 20 mm, respectively. The ROC curve (Fig. 3) analysis for predicting N1b established the optimal tumour volume cut-off value to be 350 mm3 (sensitivity= 41.6%, specificity= 78.26%) with the area under the curve being 0.587 (P = 0.019). Using the exponential equation from Fig. 2, a tumour volume of 350 mm3 corresponds to a maximum diameter measurement of 12.3 mm.
Recategorizing tumours and association of LVI with small and large volume groups
The 350 mm3 cut-off was used to classify two volume groups – a larger volume group of tumours > 350 mm3 and a smaller volume group of tumours ≤ 350 mm3. A total of 39 (11%) tumours that were >10 mm in greatest diameter were recategorized into the smaller volume group, while 10 (6%) micro-DTC tumours were recategorized into the larger volume group. This resulted in 199 patients (38.0%) in the small tumour volume group and 324 patients (62.0%) in the large tumour volume group.
Assessing the various tumour risk factors and clinicopathological characteristics, both volume and greatest diameter had statistical differences between sizes for LVI and the presence of five or more positive lymph nodes (Table 2). The greatest diameter cut-off (>10 mm) was not able to predict for presence of N1b lymph node disease, whereas tumour volume group did (P = 0.003 vs 0.097). Conversely, greatest diameter (≤10 mm) could predict for minor extrathyroidal extension (P = 0.036 vs 0.056).
Association between clinicopathological characteristics using micro-DTC cut-off and tumour volume groups created.
Clinicopathological characteristic | Using micro-DTC cut-off | Using tumour volume groups created | ||||
---|---|---|---|---|---|---|
Greatest diameter | P value | Small tumour volume group (n = 199) | Large tumour volume group (n = 324) | P value | ||
≤10 mm (n = 170) | >10 mm (n = 353) | |||||
Lymphovascular invasion | ||||||
No | 136 (80.0%) | 249 (70.5%) | 0.026 | 163 (81.9%) | 222 (68.5%) | 0.007 |
Yes | 34 (20.0%) | 104 (29.5%) | 36 (18.1%) | 102 (31.5%) | ||
Total positive lymph nodes | ||||||
≤5 | 163 (95.9%) | 301 (85.3%) | <0.001 | 190 (95.5%) | 274 (84.6%) | <0.001 |
>5 | 7 (4.1%) | 52 (14.7%) | 9 (4.5%) | 50 (15.4%) | ||
Lateral lymph node disease (N1b) | ||||||
No | 154 (90.6%) | 300 (85.0%) | 0.097 | 184 (92.5%) | 270 (83.3%) | 0.003 |
Yes | 16 (9.4%) | 53 (15.0%) | 15 (7.5%) | 54 (16.7%) | ||
Minor extrathyroidal extension | ||||||
No | 145 (85.3%) | 273 (77.3%) | 0.036 | 168 (84.4%) | 250 (77.2%) | 0.056 |
Yes | 25 (14.7%) | 80 (22.7%) | 31 (15.6%) | 74 (22.8%) |
micro-DTC, micro-differentiated thyroid cancers.
The univariate and multivariate analyses for predicting the presence of LVI are shown in Table 3 (A) and (B) using volume and greatest diameter, respectively. In the univariate analysis, the presence of more than five metastatic lymph nodes (OR = 5.555, P = <0.001) and minor ETE (OR = 2.358, P = <0.001) were both associated with LVI. Both greatest diameter > 10 mm (OR = 1.671, P = 0.022) and volume > 350 mm3 (OR = 2.080, P = 0.001) were significant in the univariate analysis. However, in the multivariate analysis, only volume > 350 mm3 remained significant as an independent predictor of LVI (OR = 1.7, P = 0.02).
Univariate and multivariate regression analysis for prediction of LVI.
Variable | Univariate analysis | Multivariate analysis | ||
---|---|---|---|---|
Odds ratio (95% CI) | P-value | Odds ratio (95% CI) | P-value | |
(A) Lymphovascular invasion and tumour volume | ||||
Female sex | 0.814 (0.504–1.315) | 0.400 | ||
Age > 55 | 0.725 (0.480–1.096) | 0.127 | ||
Hashimoto’s disease | 0.996 (0.471–2.108) | 0.992 | ||
>5 Involved lymph nodes | 5.555 (3.150–9.798) | <0.001 | 4.563 (2.546–8.178) | <0.001 |
Minor extrathyroidal extension | 2.358 (1.502–3.703) | <0.001 | 1.993 (1.237–3.210) | 0.005 |
Tumour volume > 350 mm3 | 2.080 (1.353–3.199) | 0.001 | 1.678 (1.072–2.627) | 0.024 |
Multifocality | 1.112 (0.753–1.641) | 0.595 | ||
(B) Lymphovascular invasion and greatest diameter | ||||
Female sex | 0.814 (0.504–1.315) | 0.400 | ||
Age > 55 | 0.725 (0.480–1.096) | 0.127 | ||
Hashimoto’s disease | 0.996 (0.471–2.108) | 0.992 | ||
>5 Involved lymph nodes | 5.555 (3.150–9.798) | <0.001 | 4.807 (2.685–8.607) | 0.004 |
Minor extrathyroidal extension | 2.358 (1.502–3.703) | <0.001 | 2.017 (1.254–3.244) | <0.001 |
Greatest diameter > 10 mm | 1.671 (1.076–2.594) | 0.022 | 1.314 (0.830–2.080) | 0.244 |
Multifocality | 1.112 (0.753–1.641) | 0.595 |
Analysis of patients without lateral nodal disease
By using the volume cut-off of 350 mm3, a separate analysis was conducted on the same cohort after excluding all N1b patients (n = 69) which left a cohort of 454 patients.
The multivariate analysis for LVI confirmed that the volume cut-off of 350 mm3 remained a significant predictor (OR = 1.840, P = 0.029) while the larger greatest diameter (P = 0.133) group was not significant (Table 4A andB).
Univariate and multivariate regression analyses for prediction of LVI excluding N1b patients.
Variable | Univariate analysis | Multivariate analysis | ||
---|---|---|---|---|
Odds ratio (95% CI) | P-value | Odds ratio (95% CI) | P-value | |
(A) Lymphovascular invasion and tumour volume | ||||
Female sex | 1.043 (0.572–1.905) | 0.890 | ||
Age > 55 | 0.845 (0.521–1.361) | 0.489 | ||
Hashimoto’s disease | 0.414 (0.123–1.393) | 0.154 | ||
>5 Involved lymph nodes | 2.634 (1.057–6.563) | 0.038 | 2.312 (0.912–5.860) | 0.077 |
Minor extrathyroidal extension | 2.161 (1.255–3.720) | 0.005 | 2.095 (1.209–3.631) | 0.008 |
Tumour volume > 350 mm3 | 1.944 (1.132–3.341) | 0.016 | 1.840 (1.064–3.183) | 0.029 |
Multifocality | 1.103 (0.695–1.752) | 0.676 | ||
(B) Lymphovascular invasion and greatest diameter | ||||
Female sex | 1.043 (0.572–1.905) | 0.890 | ||
Age > 55 | 0.845 (0.521–1.361) | 0.489 | ||
Hashimoto’s disease | 0.414 (0.123–1.393) | 0.154 | ||
>5 Involved lymph nodes | 2.634 (1.057–6.563) | 0.038 | 2.277 (0.894–5.802) | 0.085 |
Minor extrathyroidal extension | 2.161 (1.255–3.720) | 0.005 | 2.094 (1.210–3.623) | 0.008 |
Tumour diameter > 10 mm | 1.638 (1.974–2.754) | 0.063 | 1.501 (0.884–2.550) | 0.133 |
Multifocality | 1.103 (0.695–1.752) | 0.676 |
Discussion
With smaller DTCs having a generally favourable prognosis, treatment has been trending towards more passive approaches such as AS and hemithyroidectomy (Haugen et al. 2016, Russ et al. 2017) to minimise surgical morbidity. This study looked at whether tumour volume could be a better indicator for determining if a patient with a small DTC is a candidate for less aggressive treatment. While macroscopic lymph node involvement can be assessed using ultrasound pre-operatively, LVI is only able to be determined post-operatively on histopathological examination (DiMarco et al. 2019). Thus, utilising a factor that may predict LVI as an endpoint may be more useful for pre-operative work-up of a patient.
This study examined whether tumour volume is a better predictor of LVI than traditional size measurements. The study cohort was 81% female which is in line with the typical approximate 4:1 female to male ratio reported in other endocrine surgery studies (Sanders & Cady 1998, LeClair et al. 2021). The 93.3% proportion of PTCs in this cohort is comparable to the literature (Yip & Sosa 2016, Grani et al. 2018).
This study shows that as the greatest tumour dimension increases, there is increasing variation in tumour volume. A study by Pennington et al. used sonographic volumetric assessment and found similar findings for their larger tumours – true tumour size or volume was being misrepresented by taking a unidimensional greatest dimension measurement (Pennington et al. 2018). While Pennington et al. used pre-operative sonography data and this study used pathological data, pre-operative thyroid sonography has been shown to correlate well and have a very high positive predictive value agreement with post-operative histopathological assessment (Park et al. 2009, Moon et al. 2011) in terms of greatest diameter.
Our regression analysis showed that the relationship between volume and greatest dimension was better explained by an exponential rather than a linear model. Our results show there were 10 mm greatest dimension tumours that had a greater volume than many than 11, 12 and even some 20 mm greatest dimension tumours (Fig. 2). This could explain previous research that tumour staging with greatest diameter is not correlated precisely with true tumour burden (Ball et al. 2006).
ROC curve analysis showed a tumour volume of 350 mm3 was the optimal cut-off in predicting N1b disease. By creating a cut-off, all ≤20 mm greatest diameter tumours were able to be separated into two volume groups, and this was compared with the current DTC T1a and T1b staging which uses the greatest diameter = 10 mm cut-off by AJCC (Edge 2017). The study by Park et al. on tumour volume as a predictor for occult central lymph node metastasis in PTC also performed a ROC analysis to determine the optimal tumour volume cut-off (Park et al. 2015). This was found to be 385 mm3 which is close to what was established in our cut-off.
The tumour volume cut-off point of 350 mm3 corresponds to greatest diameter of 12.3 mm which is slightly greater than the current cut-off for micro-DTC of 10 mm. To our knowledge, there have not been any studies that investigated the relationship between volume and LVI. Tumour volume has shown promise in AS studies as an early indicator for cancer growth kinetics (Tuttle et al. 2017). Furthermore, Lim et al. (2017) found that pre-operative tumour volume measurements can be more useful than greatest diameter for predicting the risk of recurrence. However, the aforementioned studies excluded patients with higher risk pathological features such as LVI or ETE (Lim et al. 2017, Tuttle et al. 2017).
Our rate of LVI was 26.4% which lies between previous studies’ incidence rates from different institutes, with a varying incidence of LVI of 3–47% (Nishida et al. 2002, Falvo et al. 2005, Mete & Asa 2011, Wreesmann et al. 2015). Our results indicated that a pathological tumour volume of >350 mm3 was an independent predictor of LVI on multivariate analysis. The other predictive factors for LVI were greater than five metastatic lymph nodes and minor ETE, which concurs with previous studies and suggests that LVI typically occurs in combination with other clinicopathological indicators (Wreesmann et al. 2015). Interestingly, the age cut-off did not have a statistically significant correlation with LVI, in contrast to previous studies (Can et al. 2015, Sezer et al. 2017).
Further to these findings, only tumour volume was a significant predictor for LVI after excluding N1b patients. This is significant as N1b was used to create the optimal cut-off for tumour volume; hence showing the cut-off still holds without this subset of patients suggests it still holds utility. In addition, N1b patients may be able to be identified pre-operatively and, posing a greater risk, would not be suitable for a passive treatment approach.
There are some limitations to this study; since all three dimensions are not routinely reported in final histopathological analysis, there may be some selection bias. This project also optimised the volume cut-off in terms of N1b which may also be a source of bias. However, this project was intended as a pilot study to support further research to investigate the relationship between pre-operative radiological volume measurements with pathological risk stratification. Finally, we used the outcome of LVI as a surrogate endpoint, and future research should investigate the relationship of volume to clinical outcomes such as recurrence.
Conclusion
This study showed that there is a wide range of tumour volumes associated with a single measure of greatest dimension and that tumour volume is a better predictor of LVI. The results suggest that volume has the potential to be used as an indicator for factors that cannot yet be reliably predicted or determined pre-operatively such as LVI. Since volume can be obtained pre-operatively by radiological assessment, routine volume measurements may allow clinicians to perform a more accurate preoperative risk assessment and facilitate clinical management decision.
Supplementary materials
This is linked to the online version of the paper at https://doi.org/10.1530/EO-22-0066.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
Anthony Glover is supported by a Cancer Institute NSW Early Career Fellowship (2019/ECF1081) for this project.
Acknowledgements
The authors wish to acknowledge Stan Sidhu, Leigh Delbridge and the RNSH Department of Pathology for contributing to the database used for this study.
References
Ball DL, Fisher R, Burmeister B, Graham P, Joseph D, Penniment M, Krawitz H, Wheeler G, Poulsen M & Vinod S et al.2006 Stage is not a reliable indicator of tumor volume in non-small cell lung cancer: a preliminary analysis of the Trans-Tasman Radiation Oncology Group 99-05 database. Journal of Thoracic Oncology 1 667–672. (https://doi.org/10.1016/S1556-0864(1530379-8)
Boucai L, Bernet V, Shaha A, Shindo ML, Stack BC & Tuttle RM 2017 Surgical considerations for papillary thyroid microcarcinomas. Journal of Surgical Oncology 116 269–274. (https://doi.org/10.1002/jso.24673)
Can N, Tastekin E, Ozyilmaz F, Sezer YA, Guldiken S, Sut N, Sarikas N, Oz Puyan F, Guler B & Ayturk S et al.2015 Histopathological evidence of lymph node metastasis in papillary thyroid carcinoma. Endocrine Pathology 26 218–228. (https://doi.org/10.1007/s12022-015-9382-7)
Dal Maso L, Panato C, Franceschi S, Serraino D, Buzzoni C, Busco S, Ferretti S, Torrisi A, Falcini F & Zorzi M et al.2018 The impact of overdiagnosis on thyroid cancer epidemic in Italy, 1998–2012. European Journal of Cancer 94 6–15. (https://doi.org/10.1016/j.ejca.2018.01.083)
Davies L & Hoang JK 2021 Thyroid cancer in the USA: current trends and outstanding questions. Lancet: Diabetes and Endocrinology 9 11–12. (https://doi.org/10.1016/S2213-8587(2030372-7)
Davies L & Welch HG 2014 Current thyroid cancer trends in the United States. JAMA Otolaryngology: Head and Neck Surgery 140 317–322. (https://doi.org/10.1001/jamaoto.2014.1)
DiMarco AN, Wong MS, Jayasekara J, Cole-Clark D, Aniss A, Glover AR, Delbridge LW, Sywak MS & Sidhu SB 2019 Risk of needing completion thyroidectomy for low-risk papillary thyroid cancers treated by lobectomy. BJS Open 3 299–304. (https://doi.org/10.1002/bjs5.50137)
Edge SB 2017 AJCC Cancer Staging Manual, 8th ed. New York, NY, USA: Springer.
Falvo L, Catania A, D’andrea V, Marzullo A, Giustiniani MC & De Antoni E 2005 Prognostic importance of histologic vascular invasion in papillary thyroid carcinoma. Annals of Surgery 241 640–646. (https://doi.org/10.1097/01.sla.0000157317.60536.08)
Grani G, Lamartina L, Durante C, Filetti S & Cooper DS 2018 Follicular thyroid cancer and Hürthle cell carcinoma: challenges in diagnosis, treatment, and clinical management. Lancet: Diabetes and Endocrinology 6 500–514. (https://doi.org/10.1016/S2213-8587(1730325-X)
Guth S, Theune U, Aberle J, Galach A & Bamberger CM 2009 Very high prevalence of thyroid nodules detected by high frequency (13 MHz) ultrasound examination. European Journal of Clinical Investigation 39 699–706. (https://doi.org/10.1111/j.1365-2362.2009.02162.x)
Haugen BR, Alexander EK, Bible KC, Doherty GM, Mandel SJ, Nikiforov YE, Pacini F, Randolph GW, Sawka AM & Schlumberger M et al.2016 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 and differentiated thyroid cancer. Thyroid 26 1–133. (https://doi.org/10.1089/thy.2015.0020)
Ito Y, Miyauchi A, Kihara M, Fukushima M, Higashiyama T & Miya A 2018 Overall survival of papillary thyroid carcinoma patients: a single-institution long-term follow-up of 5897 patients. World Journal of Surgery 42 615–622. (https://doi.org/10.1007/s00268-018-4479-z)
Jiang Y, You K, Qiu X, Bi Z, Mo H, Li L & Liu Y 2018 Tumor volume predicts local recurrence in early rectal cancer treated with radical resection: a retrospective observational study of 270 patients. International Journal of Surgery 49 68–73. (https://doi.org/10.1016/j.ijsu.2017.11.052)
LeClair K, Bell KJL, Furuya-Kanamori L, Doi SA, Francis DO & Davies L 2021 Evaluation of gender inequity in thyroid cancer diagnosis: differences by sex in US thyroid cancer incidence compared with a meta-analysis of subclinical thyroid cancer rates at autopsy. JAMA Internal Medicine 181 1351–1358. (https://doi.org/10.1001/jamainternmed.2021.4804)
Lim ST, Jeon YW & Suh YJ 2017 The prognostic values of preoperative tumor volume and tumor diameter in T1N0 papillary thyroid cancer. Cancer Research and Treatment 49 890–897. (https://doi.org/10.4143/crt.2016.325)
Mete O & Asa SL 2011 Pathological definition and clinical significance of vascular invasion in thyroid carcinomas of follicular epithelial derivation. Modern Pathology 24 1545–1552. (https://doi.org/10.1038/modpathol.2011.119)
Moon HJ, Yoon JH, Kwak JY, Chung WY, Nam KH, Jeong JJ & Kim EK 2011 Positive predictive value and interobserver variability of preoperative staging sonography for thyroid carcinoma. American Journal of Roentgenology 197 W324–W330. (https://doi.org/10.2214/AJR.10.5576)
Nishida T, Katayama Si & Tsujimoto M 2002 The clinicopathological significance of histologic vascular invasion in differentiated thyroid carcinoma. American Journal of Surgery 183 80–86. (https://doi.org/10.1016/s0002-9610(0100843-1)
Park JS, Son KR, Na DG, Kim E & Kim S 2009 Performance of preoperative sonographic staging of papillary thyroid carcinoma based on the sixth edition of the AJCC/UICC TNM classification system. American Journal of Roentgenology 192 66–72. (https://doi.org/10.2214/AJR.07.3731)
Park KN, Kang KY, Hong HS, Jeong HS & Lee SW 2015 Predictive value of estimated tumor volume measured by ultrasonography for occult central lymph node metastasis in papillary thyroid carcinoma. Ultrasound in Medicine and Biology 41 2849–2854. (https://doi.org/10.1016/j.ultrasmedbio.2015.02.018)
Park SY, Kim HI, Kim JH, Kim JS, Oh YL, Kim SW, Chung JH, Jang HW & Kim TH 2018 Prognostic significance of gross extrathyroidal extension invading only strap muscles in differentiated thyroid carcinoma. British Journal of Surgery 105 1155–1162. (https://doi.org/10.1002/bjs.10830)
Pennington TE, Thwin M, Sywak M, Delbridge L & Sidhu S 2018 Sonographic volumetric assessment is a more accurate measure than maximum diameter alone in papillary thyroid cancer. Journal of the Endocrine Society 2 1284–1292. (https://doi.org/10.1210/js.2018-00233)
Russ G, Bonnema SJ, Erdogan MF, Durante C, Ngu R & Leenhardt L 2017 European Thyroid Association guidelines for ultrasound malignancy risk stratification of thyroid nodules in adults: the EU-TIRADS. European Thyroid Journal 6 225–237. (https://doi.org/10.1159/000478927)
Sanders LE & Cady B 1998 Differentiated thyroid cancer: reexamination of risk groups and outcome of treatment. Archives of Surgery 133 419–425. (https://doi.org/10.1001/archsurg.133.4.419)
Sezer A, Celik M, Yilmaz Bulbul B, Can N, Tastekin E, Ayturk S, Ustun F, Guldiken S & Sut N 2017 Relationship between lymphovascular invasion and clinicopathological features of papillary thyroid carcinoma. Bosnian Journal of Basic Medical Sciences 17 144–151. (https://doi.org/10.17305/bjbms.2017.1924)
Stamey TA, Mcneal JE, Yemoto CM, Sigal BM & Johnstone IM 1999 Biological determinants of cancer progression in men with prostate cancer. JAMA 281 1395–1400. (https://doi.org/10.1001/jama.281.15.1395)
Su XD, Xie HJ, Liu QW, Mo YX, Long H & Rong TH 2017 The prognostic impact of tumor volume on stage I non-small cell lung cancer. Lung Cancer 104 91–97. (https://doi.org/10.1016/j.lungcan.2016.12.013)
Takenaka T, Yamazaki K, Miura N, Mori R & Takeo S 2016 The prognostic impact of tumor volume in patients with clinical stage IA non-small cell lung cancer. Journal of Thoracic Oncology 11 1074–1080. (https://doi.org/10.1016/j.jtho.2016.02.005)
Tuttle RM 2018 Controversial issues in thyroid cancer management. Journal of Nuclear Medicine 59 1187–1194. (https://doi.org/10.2967/jnumed.117.192559)
Tuttle RM, Fagin JA, Minkowitz G, Wong RJ, Roman B, Patel S, Untch B, Ganly I, Shaha AR & Shah JP et al.2017 Natural history and tumor volume kinetics of papillary thyroid cancers during active surveillance. JAMA Otolaryngology: Head and Neck Surgery 143 1015–1020. (https://doi.org/10.1001/jamaoto.2017.1442)
Unal I 2017 Defining an optimal cut-point value in ROC analysis: an alternative approach. Computational and Mathematical Methods in Medicine 2017 3762651. (https://doi.org/10.1155/2017/3762651)
Wreesmann VB, Nixon IJ, Rivera M, Katabi N, Palmer F, Ganly I, Shaha AR, Tuttle RM, Shah JP & Patel SG et al.2015 Prognostic value of vascular invasion in well-differentiated papillary thyroid carcinoma. Thyroid 25 503–508. (https://doi.org/10.1089/thy.2015.0052)
Xiao W, Liu S, Tian Y, Guan Y, Huang S, Lin C, Zhao C, Lu T & Han F 2015 Prognostic significance of tumor volume in locally recurrent nasopharyngeal carcinoma treated with salvage intensity-modulated radiotherapy. PLoS ONE 10 e0125351. (https://doi.org/10.1371/journal.pone.0125351)
Yip L & Sosa JA 2016 Molecular-directed treatment of differentiated thyroid cancer: advances in diagnosis and treatment. JAMA Surgery 151 663–670. (https://doi.org/10.1001/jamasurg.2016.0825)
Zanocco KA, Hershman JM & Leung AM 2019 Active surveillance of low-risk thyroid cancer. JAMA 321 2020–2021. (https://doi.org/10.1001/jama.2019.5350)