Estimating the Growth Rate of Lung Metastases in Differentiated Thyroid Carcinoma: Response Evaluation Criteria in Solid Tumors or Doubling Time?

Abstract

Background:

Estimating the growth rate of lung metastases for the treatment of patients with metastases of differentiated thyroid carcinoma (DTC) is important. This study aimed to evaluate survival outcomes according to different criteria for estimating the growth rate of lung metastases.

Methods:

Patients with macronodular (≥1 cm) lung metastases of DTC who underwent total thyroidectomy and high-dose radioactive iodine therapy between 1995 and 2013 were enrolled. The time to progressive disease (PD) by the Response Evaluation Criteria in Solid Tumors (RECIST), average tumor volume doubling time of the two dominant target lung lesions (midDT), and thyroglobulin doubling time (TgDT) were measured in each patient, and their association with disease-specific survival (DSS) was evaluated.

Results:

Forty-four patients with target lung metastatic nodules with an initial maximal diameter of 1.3 cm (median) were followed-up for a median of 6.8 years after the diagnosis of lung metastases. Based on RECIST, 12 patients (27.3%) showed fast tumor progression, with time to PD <1 year. When assessed by midDT, nine patients (20.5%) had midDT ≤1 year, showing rapid tumor progression. Seven of 33 patients (21.2%) who were negative for thyroglobulin antibody had midDT <1 year. Growth rates assessed by all three criteria were significantly associated with DSS. However, midDT had the highest predictive value for DSS, with a proportion of variation explained of 33.6%. Five-year DSS was 29.6% in patients with midDT ≤1 year, 50.0% in patients with time to PD <1 year, and 42.9% in patients with TgDT <1 year.

Conclusions:

Among the different criteria for estimating the growth rate of metastases in patients with lung metastases of DTC, midDT was the most powerful for predicting DSS, in comparison with RECIST and TgDT. Performing at least three serial chest computed tomography scans during the first year from the diagnosis of lung metastases can facilitate early detection of patients with rapid tumor progression and provide objective guidance for initiation of systemic therapy.

Introduction

Sorafenib and lenvatinib, which are oral multikinase inhibitors mainly targeting angiogenesis (1,2), have been approved for treating patients with radioactive iodine (RAI)-refractory differentiated thyroid carcinoma (DTC), with promising results. In phase 3 trials, sorafenib and lenvatinib prolonged progression-free survival by a median of 5 months (3) and 14.7 months (4), respectively, compared with placebo. These two tyrosine kinase inhibitors (TKIs) are now widely used in real-world settings, with very encouraging outcomes (5 –9). However, despite these results, neither sorafenib nor lenvatinib is indicated for all patients with RAI-refractory DTC. Clinicians pay close attention when initiating these agents, as severe toxicities are frequently noted. Adverse events of grade ≥3 occurred in 20.3% of patients treated with sorafenib (3) and 75.9% of patients treated with lenvatinib (4); in addition, treatment-attributed deaths occurred in 2.3% of patients treated with lenvatinib. Thus, it is important to identify which patients would actually benefit from treatment with TKIs and when would be the appropriate time to start TKIs.

The 2015 American Thyroid Association guidelines provide three factors favoring the initiation of TKIs (10): symptomatic disease; diffuse disease progression; and imminently threatening disease progression, with a high likelihood of requiring intervention and/or leading to morbidity or mortality in <6 months. Thus, estimating the tumor growth rate using an objective standard to select patients with “threatening disease progression” is crucial. Schlumberger et al. suggested the use of the Response Evaluation Criteria in Solid Tumors (RECIST) (11) to define disease progression and proposed initiation of TKI therapy when disease progression evaluated by RECIST is observed within 12–14 months (12). The usefulness of RECIST for predicting survival outcomes in patients with macronodular lung metastases was demonstrated in our previous study (13). We observed that disease progression within the first year according to RECIST was an independent predictor of cancer-specific survival and that patients in this condition are eligible for treatment with TKIs.

More recently, Sabra et al. assessed tumor volume doubling time (TVDT) as a means of measuring disease progression in patients with lung metastases (14). They reported that average tumor volume doubling time of the two dominant target lung lesions (midDT) is a good predictor of overall survival (OS) and that patients with midDT <1 year would be the best candidates for TKI therapy. This concept of midDT was based on the thyroglobulin doubling time (TgDT) suggested by Miyauchi et al. in 2011 (15). They calculated the TgDT in patients negative for thyroglobulin antibody (TgAb) to estimate tumor burden and demonstrated that TgDT is a robust predictor of disease-specific survival (DSS).

Currently, there are three criteria for estimating the rate of tumor growth: RECIST, midDT, and TgDT. To the best of our knowledge, no previous studies have compared the predictability of survival according to these three criteria or assessed their correlation with each other. The aim of this study was to evaluate the tumor growth rate and compare the clinical outcomes using different criteria in patients with lung metastases enrolled in our previous study (13). The findings of this study may be used to provide the optimal method to identify patients eligible for TKI treatment.

Materials and Methods

Study design and patients

This study is a retrospective cohort study of patients with macronodular (≥1 cm) lung metastases of DTC, as described in our previous study (13). Patients who underwent total thyroidectomy with subsequent high-dose RAI treatment between 1995 and 2013 at the Asan Medical Center in Korea were selected. The exclusion criteria were age <18 years and inadequate follow-up computed tomography (CT) scans. A total of 44 patients were included in this study. Treatment and follow-up protocol were as described in the previous study (13). Data collection and subsequent analysis were approved by the institutional review board of the Asan Medical Center.

Measurement of lung metastases and thyroglobulin

Up to two largest metastatic lung nodules with diameters >1 cm were selected as the target lesions regardless of their RAI avidity. The size of the target lung nodules was measured on serial chest CT scans by two experienced radiologists (S.M.L. and J.H.L.). For patients who initiated TKI treatment, the lesions were measured until the initiation of treatment. The scanners used for CT imaging were Somatom Definition, Sensation-16 (Siemens Medical Systems, Forchheim, Germany), Lightspeed Volume CT, and Discovery HD 750 (GE Healthcare, Milwaukee, WI)—with 120 kVp, 30–200 mAs. Images were reconstructed using the sharp reconstruction algorithm with thicknesses of 2.5–3.0 mm.

Serum thyroglobulin (Tg) levels were measured with the Tg-plus RIA kit (Brahms AG, Hennigsdorf, Germany) with a functional sensitivity (20% interassay variation coefficient) of 0.2 ng/mL and an analytical sensitivity of 0.08 ng/mL (16). Serum TgAb levels were measured with the anti-Tg RIA kit (Brahms AG) with a functional sensitivity (20% interassay variation coefficient) of 0.07 ng/mL (17). A value of ≥60 IU/mL was designated as the minimum threshold denoting TgAb positivity.

Evaluation of tumor growth rate

Time to progressive disease by RECIST

Disease progression of the lung metastases was evaluated according to RECIST v1.1 (11). The sum of the longest diameters of all (one or two) target lung lesions from the first CT scan was defined as the baseline tumor size. With this baseline tumor size as reference, disease responses were defined as follows: complete response (no visible lesions in radiological images); partial response (at least 30% decrease in the sum of the diameters of the target lesions); progressive disease (PD; at least 20% increase in the sum of the diameters of the target lesions, appearance of new lesions, increase in size of one or more nontarget lesions, or any increase in the size of nontarget lesions resulting in increased overall tumor burden); and stable disease (SD; neither sufficient shrinkage to qualify for partial response nor sufficient increase to qualify for PD).

Tumor volume doubling time

midDT was calculated by using the calculator provided by Kuma Hospital (www.kuma-h.or.jp/english/about/doubling-time-progression-calculator). At least three serial maximal diameters were used for the calculation of midDT. As described by Sabra et al. (14), measurement of midDT uses a nonlinear square regression to compute a regression line, log y = log a + bx, with x defined as the time interval from the initial CT scan and y as the tumor volume. midDT is calculated as (log 2)/b. For patients with two target lesions, the average of the midDT values calculated in each nodule was used.

Tg doubling time

Serum Tg levels in patients receiving L-thyroxine suppression therapy measured within 3 months of the time of measurement of the lung nodules were used to calculate TgDT. TgDT was also calculated by using the calculator provided by Kuma Hospital. Patients with positivity for TgAb (≥60 IU/mL) were excluded from the TgDT analyses.

Classification of patients

Patients were first classified into three groups according to the time to PD: patients with PD observed in less than a year after the diagnosis of lung metastases (PD <1 year), patients with PD observed between 1 and 3 years after the diagnosis of lung metastases (PD 1–3 years), and patients with PD observed 3 or more years after the diagnosis of lung metastases or with SD up to the last CT scan (PD ≥3 years).

The patients were then classified into three groups according to midDT: midDT of 1 year or less (midDT ≤1 year), midDT of 1–3 years (midDT 1–3 years), and midDT of 3 years or more or negative midDT (midDT ≥3 years).

Finally, the patients were classified into three groups according to their TgDT: TgDT of less than a year (TgDT <1 year), TgDT of 1–3 years (TgDT 1–3 years), and TgDT of 3 years or more or negative TgDT (TgDT ≥3 years).

Based on the previous studies, we used the same intervals for dividing the periods for each criterion and the intervals were arbitrarily selected by the authors.

Statistical analysis

R version 3.4.0 was used for data analysis (R Foundation for Statistical Computing, Vienna, Austria), and graphs were drawn with GraphPad Prism version 5.0 (GraphPad Software, San Diego, CA). Continuous variables are presented as medians with interquartile ranges (IQRs), and categorical variables are presented as numbers with percentages. DSS was calculated from the time of diagnosis of lung metastases until death from DTC or the last censoring, and OS was defined as the time from the diagnosis of lung metastases until death from any cause or the last censoring. Both DSS and OS curves were plotted using the Kaplan–Meier method. The log-rank test was used to determine the significance of differences. The predictability of DSS was analyzed by calculating the proportion of variation explained (PVE), as previously described (18). The formula PVE = 1 − exp(−G²/n) was used to calculate PVE (19), where G² is the maximum likelihood ratio determined by the chi-squared test and Cox regression analyses, and n is the total number of cases in the present study. PVE (%) ranges from 0 to 100, and higher values indicate better predictability of DSS. Differences with p < 0.05 (two-sided) were regarded as significant.

Results

Baseline characteristics

Table 1 presents the baseline clinicopathologic characteristics of the 44 patients. About two-thirds of the patients were diagnosed with papillary thyroid carcinoma, while the other third were diagnosed with follicular thyroid carcinoma. Patients with Hürthle cell carcinoma or poorly differentiated carcinoma were not eligible. The median age of patients at the time of diagnosis of lung metastases was 61 years; 55% patients were female. The median primary tumor size was 3.3 cm, and the median initial largest diameter of the target lung nodule was 1.3 cm. Of the 44 patients, 24 had only lung metastasis, 17 additionally had bone metastasis, and 3 had lung and brain metastases. The median initial serum Tg level was 47.4 ng/mL, which increased to a median of 87.5 ng/mL at the last measurement. Among the 33 patients negative for TgAb, TgDT <1 year was observed in 7 patients (21.2%), TgDT 1–3 years in 13 patients (39.4%), and TgDT ≥3 years in 13 patients (39.4%).

Table 1.

Baseline Characteristics of the Study Patients

Characteristic	Value ^a (N = 44)
Pathology
PTC	31 (70%)
FTC	13 (30%)
Age (years)	61 (52–66)
Sex (female)	24 (55.0%)
Primary tumor size (cm)	3.3 (2.5–4.7)
Extrathyroidal extension	31 (70%)
Lymph node metastases	31 (70%)
Initial maximal diameter of target lung nodules (cm)	1.3 (1.0–1.8)
RAI avidity	18 (41%)
Metastatic sites at distant metastasis diagnosis
Lung only	24 (54.5%)
Lung and bone	17 (38.6%)
Lung and brain	3 (6.8%)
Initial serum Tg (ng/mL)^b	47.4 (12.0–401.0)
Last serum Tg (ng/mL)^b	87.5 (13.5–631.5)
TgDT^b
<1 Year	7 (21.2%, 7/33)
1–3 Years	13 (39.4%, 13/33)
≥3 Years	13 (39.4%, 13/33)
Follow-up duration after lung metastases diagnosis (years)	6.8 (5.3–10.4)

Values are number (percentage) or median (interquartile range).

Patients who were positive for TgAb were excluded.

FTC, follicular thyroid carcinoma; PTC, papillary thyroid carcinoma; RAI, radioactive iodine; Tg, thyroglobulin; TgAb, thyroglobulin antibody; TgDT, thyroglobulin doubling time.

Distribution of patients

As shown in Table 2, 12 patients (27.3%) were classified in the PD <1 year group, 17 patients (38.6%) in the PD 1–3 years group, and 15 patients (34.1%) in the PD ≥3 years group. There were 9 patients (20.5%) in the midDT <1 year group, 20 patients (45.4%) in the midDT 1–3 years group, and 15 patients (34.1%) in the midDT ≥3 years group. Time to PD and midDT were strongly correlated (Pearson r = 0.82, p < 0.001).

Table 2.

Distribution of Patients According to Time to Progressive Disease and Tumor Volume Doubling Time

Time to PD	midDT
Time to PD	< 1 Year	1–3 Years	≥3 Years	Total
<1 Year	8	3	1	12
1–3 Years	1	15	1	17
≥3 Years	0	2	13	15
Total	9	20	15	44

midDT, average tumor volume doubling time; PD, progressive disease.

Survival outcomes and growth rate of lung metastases

During a median follow-up period of 6.8 (IQR, 5.3–10.4) years from the time of diagnosis of lung metastases, 25 patients (56.8%) died due to progression of DTC. Figure 1 shows the DSS curves based on time to PD (Fig. 1A) and midDT (Fig. 1B). DSS was significantly different according to both criteria (log-rank p = 0.046 for time to PD and p < 0.001 for midDT). In addition, Figure 1C shows DSS according to TgDT, which also reveals significant differences among the three groups (log-rank p = 0.002). We also evaluated OS using the three criteria (Supplementary Fig. S1) and obtained results similar to DSS.

FIG. 1.

Disease-specific survival based on (A) time to PD, (B) midDT, and (C) TgDT. midDT, average tumor volume doubling time; PD, progressive disease; TgDT, thyroglobulin doubling time.

Predictability of survival outcomes based on time to PD and midDT

Although all the criteria were associated with DSS, patients with midDT ≤1 year had distinctly poorer DSS than patients with PD <1 year or TgDT <1 year (5-year DSS: 29.6% in the midDT ≤1 year group, 50.0% in the PD <1 year group, and 42.9% in the TgDT <1 year group, Table 3). midDT showed the highest predictive value for DSS (PVE 33.6%). PVE was 13.4% for RECIST and 17.8% for TgDT. Supplementary Table S1 summarizes PVE for OS, which also shows highest predictive value by midDT (PVE 40.5%).

Table 3.

Cox Regression Model and Proportion of Variation Explained of Criteria for Estimated Tumor Growth Rate

	N	5-Year DSS [95% CI]	HR [95% CI]	p	PVE (%)
Time to PD
<1 Year	12	50.0% [28.4–88.0%]	1.0 [Reference]	—	13.4
1–3 Years	17	88.2% [74.2–100%]	0.65 [0.25–1.63]	0.361
≥3 Years	15	93.3% [81.5–100%]	0.25 [0.08–0.80]	0.019
midDT
<1 Year	9	29.6% [10.0–87.5%]	1.0 [Reference]	—	33.6
1–3 Years	20	90.0% [77.7–100%]	0.08 [0.02–0.35]	<0.001
≥3 Years	15	93.3% [81.5–100%]	0.04 [0.08–0.20]	<0.001
TgDT
<1 Year	7	42.9% [18.2–100%]	1.0 [Reference]	—	17.8
1–3 Years	13	84.6% [67.1–100%]	0.19 [0.05–0.72]	0.015
≥3 Years	13	92.3% [78.9–100%]	0.11 [0.02–0.47]	0.003

CI, confidence interval; DSS, disease-specific survival; HR, hazard ratio; PVE, proportion of variation explained.

Discussion

The lungs are the most common site of distant metastases in patients with DTC (20 –22). It is well established that patients with lung metastases >1 cm have poor clinical outcomes (20,23). In addition to the initial tumor size, tumor growth rate, which varies from slow to rapid among patients, is an important predictor of survival outcomes (13,14). In this study, we evaluated tumor growth rate in patients with macronodular lung metastases using different criteria and compared the predictability of survival among criteria. The results showed that although RECIST, midDT, and TgDT can all predict DSS, midDT is the most powerful predictor of DSS, with the highest value of PVE for DSS. The 5-year DSS of patients with midDT <1 year was 29.6%, which indicated that these patients should be considered for the initiation of systemic therapy (such as TKI therapy). The PVE value for midDT measurement for risk stratification in patients to determine eligibility for TKI therapy (33.6%) was comparable to that for the American Thyroid Association risk stratification system to estimate recurrence risk (34%) (24).

In 1956, Collins et al. proposed that human tumors exhibited an exponential growth pattern (25). This exponential growth remains constant over a long period of time (26). In this respect, a small increase in tumor diameter can result in a much larger increase in tumor volume (27), which indicates that measuring only the unidimensional increase in size can lead to the underestimation of tumor progression (14). Considering all the results from this present study, it seems clear that three-dimensional tumor volume is more sensitive than unidimensional diameter for predicting tumor progression. Although at least three sequential CT images are required for assessing midDT, and only two are sufficient to assess the clinical response based on RECIST, the calculating techniques used in the two criteria are comparable in terms of complexity. In addition, RECIST was originally designed to assess the therapeutic response in prospective clinical trials and requires assessment to be performed as close as possible to the treatment initiation date (no longer than 4 weeks); however, there is no such restriction with the use of midDT. Moreover, RECIST can only be applied in patients with target lesions sized >1 cm, while midDT can be applied in patients with smaller lesions and who are in the early stages of their disease course. Thus, midDT can be efficiently used in clinical practice for evaluating patients with metastatic thyroid cancer before TKI therapy as it is a robust predictor of survival in patients with lung metastases. As CT scans can determine tumor volume with less error than ultrasonography, midDT measured using CT images can be utilized as a good clinical marker of tumor growth. However, interobserver reproducibility of bidimensional midDT measurements, especially when evaluating lesions <1 cm, has not been fully evaluated. Therefore, further studies evaluating reproducibility should be performed.

There is convincing evidence favoring the initiation of TKI therapy in patients with midDT ≤1 year (14). To avoid missing any cases with threatening disease progression, serial CT scans should be performed at least three times in the first year of diagnosing lung metastases to detect patients with midDT ≤1 year. Knowledge regarding the patient's midDT data during the first year of developing lung metastases can also help inform the optimal frequency of clinical and radiological follow-up. In this study, 13 patients were treated with TKIs. Initially, all of them started with sorafenib and four patients received salvage treatment with lenvatinib. Two patients achieved partial response, eight achieved SD, and three had PD. Among these, five patients are alive, of whom four are still receiving TKI therapy. However, the number of patients treated with TKI was too small to draw any correlation between the therapeutic response and TVDT. Moreover, TVDT was not considered at the time of initiation of TKIs in these patients, which limits the evaluation of TVDT with regard to the response to TKI.

In the present study, 8 of 12 patients with PD <1 year were categorized in the midDT ≤1 year group and the other 4 patients had midDT 1–4 years. If tumor growth is assumed to be constant, PD <1 year corresponds to midDT of almost 4 years (14,27), explaining the discordance between PD <1 year and midDT ≤1 year. However, there was one patient with midDT ≤1 year but PD 1–3 years. This case cannot be explained if tumor growth is constant over time. Although most tumors are expected to grow at a constant rate, there are some exceptions, and therefore routine follow-up CT scans are required at certain intervals. Despite some discrepancies, overall midDT and RECIST were strongly correlated. In addition, when the baseline characteristics of the patients in the three groups (time to PD <1 year and TVDT <1 year [n = 8], time to PD <1 year but TVDT of 1–4 years [n = 4], and time interval to PD of 1–3 years but TVDT <1 year [n = 1]) were compared, there were no differences in baseline factors to explain the discordance (Supplementary Table S2). When two serial CT scans are performed within the first year from the diagnosis of lung metastases, clinicians should assess whether the disease status corresponds to the criteria of PD by RECIST (PD <1 year by RECIST); if so, they should perform the next CT scan within a short period of time to calculate midDT, as rapid disease progression can be expected. If a CT scan cannot be performed for any reason and therefore midDT cannot be calculated, initiating TKI can be considered according to RECIST, which should be a good substitute for midDT based on our results. In patients with rapidly progressive thyroid cancer in whom midDT cannot be precisely measured, the clinical criteria for structural disease progression may be applicable as suggested by Sabra et al. (28).

A strength of our study is that it estimated tumor growth rate using different criteria in a single cohort. This enabled us to compare the predictability of survival among these criteria, which demonstrated that midDT was the most sensitive marker for disease progression. However, because of the retrospective design of the study, absolute superiority of any criterion cannot be demonstrated. Another limitation of the study is the small number of patients, which could result in underestimation of an association between RECIST and DSS. Moreover, there were 11 patients who were positive for TgAb, limiting evaluation of the predictive role of TgDT in these patients. However, even among the small number of patients who were negative for TgAb, TgDT showed a good correlation with DSS. Another limitation is the difference in the time interval between CT evaluation for PD and TVDT. A longer time interval was used in determining TVDT starting from the first CT that diagnosed lung metastases to the most recent follow-up CT or the last CT before initiating TKI therapy; this long interval may have induced an element of bias in terms of accuracy when comparing the two criteria.

Estimating tumor growth rate using serial CT scans can provide objective guidance in making decisions for initiating systemic therapy. Although conventional RECIST for assessing disease progression is associated with survival outcomes, midDT has been shown to exhibit the most powerful predictive value. TgDT can also be used for reference in patients who are negative for TgAb. We suggest at least two additional CT scans within the first year after the diagnosis of lung metastases for early identification of patients with rapid tumor progression who are eligible for initiation of TKI.

Footnotes

Acknowledgment

We acknowledge with gratitude to biostatistician Minkyu Han for his help and guidance regarding statistical analyses throughout this study.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This study was supported by the National Research Foundation (NRF) of Korea Research Grant (NRF-2018R1D1A1A02085365).

Supplementary Material

Supplementary Figure S1

Supplementary Table S1

Supplementary Table S2

References

Kim

. 2014. Molecular pathogenesis and targeted therapies in well-differentiated thyroid carcinoma. Endocrinol Metab (Seoul), 29:211–216.

Fala

. 2015. Lenvima (Lenvatinib), a multireceptor tyrosine kinase inhibitor, approved by the FDA for the treatment of patients with differentiated thyroid cancer. Am Health Drug Benefits, 8:176–179.

Brose

, Nutting

, Jarzab

, Elisei

, Siena

, Bastholt

, de la Fouchardiere

, Pacini

, Paschke

, Shong

, Sherman

, Smit

, Chung

, Kappeler

, Pena

, Molnar

, Schlumberger

. 2014. Sorafenib in radioactive iodine-refractory, locally advanced or metastatic differentiated thyroid cancer: a randomised, double-blind, phase 3 trial. Lancet, 384:319–328.

Schlumberger

, Tahara

, Wirth

, Robinson

, Brose

, Elisei

, Habra

, Newbold

, Shah

, Hoff

, Gianoukakis

, Kiyota

, Taylor

, Kim

, Krzyzanowska

, Dutcus

, de las Heras

, Zhu

, Sherman

. 2015. Lenvatinib versus placebo in radioiodine-refractory thyroid cancer. N Engl J Med, 372:621–630.

Berdelou

, Borget

, Godbert

, Nguyen

, Garcia

, Chougnet

, Ferru

, Buffet

, Chabre

, Huillard

, Leboulleux

, Schlumberger

. 2017. Lenvatinib for the treatment of radioiodine-refractory thyroid cancer in real-life practice. Thyroid, 28:72–78.

Kim

, Kim

, Shin

, Lim

, Kim

, Chung

, Shong

, Kim

. 2018. Tertiary care experience of sorafenib in the treatment of progressive radioiodine-refractory differentiated thyroid carcinoma: a Korean multicenter study. Thyroid, 28:340–348.

Nervo

, Gallo

, Sama

, Felicetti

, Alfano

, Migliore

, Marchisio

, Berardelli

, Arvat

, Piovesan

. 2018. Lenvatinib in advanced radioiodine-refractory thyroid cancer: a snapshot of real-life clinical practice. Anticancer Res, 38:1643–1649.

Balmelli

, Railic

, Siano

, Feuerlein

, Cathomas

, Cristina

, Guthner

, Zimmermann

, Weidner

, Pless

, Stenner

, Rothschild

. 2018. Lenvatinib in advanced radioiodine-refractory thyroid cancer: a retrospective analysis of the Swiss Lenvatinib named patient program. J Cancer, 9:250–255.

Suyama

, Fujiwara

, Takeshita

, Sueta

, Inao

, Yamamoto-Ibusuki

, Yamamoto

, Iwase

. 2017. Clinical features of lenvatinib treatment in elderly patients with advanced thyroid cancer. Mol Clin Oncol, 7:24–26.

10.

Haugen

, Alexander

, Bible

, Doherty

, Mandel

, Nikiforov

, Pacini

, Randolph

, Sawka

, Schlumberger

, Schuff

, Sherman

, Sosa

, Steward

, Tuttle

, Wartofsky

. 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.

11.

Eisenhauer

, Therasse

, Bogaerts

, Schwartz

, Sargent

, Ford

, Dancey

, Arbuck

, Gwyther

, Mooney

, Rubinstein

, Shankar

, Dodd

, Kaplan

, Lacombe

, Verweij

. 2009. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer, 45:228–247.

12.

Schlumberger

, Brose

, Elisei

, Leboulleux

, Luster

, Pitoia

, Pacini

. 2014. Definition and management of radioactive iodine-refractory differentiated thyroid cancer. Lancet Diabetes Endocrinol, 2:356–358.

13.

Kim

, Kim

, Park

, Kwon

, Jeon

, Lee

, Kim

, Shong

, Kim

. 2017. Growth kinetics of macronodular lung metastases and survival in differentiated thyroid carcinoma. Thyroid, 27:915–922.

14.

Sabra

, Sherman

, Tuttle

. 2017. Tumor volume doubling time of pulmonary metastases predicts overall survival and can guide the initiation of multikinase inhibitor therapy in patients with metastatic, follicular cell-derived thyroid carcinoma. Cancer, 123:2955–2964.

15.

Miyauchi

, Kudo

, Miya

, Kobayashi

, Ito

, Takamura

, Higashiyama

, Fukushima

, Kihara

, Inoue

, Tomoda

, Yabuta

, Masuoka

. 2011. Prognostic impact of serum thyroglobulin doubling-time under thyrotropin suppression in patients with papillary thyroid carcinoma who underwent total thyroidectomy. Thyroid, 21:707–716.

16.

Kim

, Jeon

, Kim

, Lee

, Ryu

, Cho

, Ko

, Lee

, Chun

, Min

, Kim

, Shong

, Kim

. 2016. Comparison of thyroglobulin measurements using three different immunoassay kits: a BRAMHS Tg-plus RIA kit, a BRAMHS hTg sensitive Kryptor kit, and a Beckman Coulter ACCESS immunoassay kit. Endocrinol Metab (Seoul), 31:462–468.

17.

Song

, Oh

, Jeon

, Chung

, Hong

, Ryu

, Baek

, Lee

, Kim

, Shong

, Kim

. 2019. The value of preoperative antithyroidperoxidase antibody as a novel predictor of recurrence in papillary thyroid carcinoma. Int J Cancer, 144:1414–1420.

18.

Song

, Lee

, Oh

, Jeon

, Song

, Kim

, Shong

, Sung

, Kim

. 2019. A relook at the T stage of differentiated thyroid carcinoma with a focus on gross extrathyroidal extension. Thyroid, 29:202–208.

19.

Brierley

, Panzarella

, Tsang

, Gospodarowicz

, O'Sullivan

. 1997. A comparison of different staging systems predictability of patient outcome: thyroid carcinoma as an example. Cancer, 79:2414–2423.

20.

Sabra

, Ghossein

, Tuttle

. 2016. Time course and predictors of structural disease progression in pulmonary metastases arising from follicular cell-derived thyroid cancer. Thyroid, 26:518–524.

21.

Kim

, Kim

, Jung

, Jeon

, Kim

, Kang

, Han

, Cho

, Kim

, Chung

. 2018. Prognosis of differentiated thyroid carcinoma with initial distant metastasis: a multicenter study in Korea. Endocrinol Metab (Seoul), 33:287–295.

22.

Cho

, Choi

, Yeom

, Lim

, Moon

, Park

, Chung

, Cho

, Yi

, Park

. 2014. Long-term prognosis of differentiated thyroid cancer with lung metastasis in Korea and its prognostic factors. Thyroid, 24:277–286.

23.

Kim

, Kim

, Park

, Kwon

, Jeon

, Lee

, Ryu

, Kim

, Shong

, Kim

. 2017. Initial size of metastatic lesions is best prognostic factor in patients with metastatic differentiated thyroid carcinoma confined to the lung. Thyroid, 27:49–58.

24.

Tuttle

, Tala

, Shah

, Leboeuf

, Ghossein

, Gonen

, Brokhin

, Omry

, Fagin

, Shaha

. 2010. Estimating risk of recurrence in differentiated thyroid cancer after total thyroidectomy and radioactive iodine remnant ablation: using response to therapy variables to modify the initial risk estimates predicted by the new American Thyroid Association staging system. Thyroid, 20:1341–1349.

25.

Collins

, Loeffler

, Tivey

. 1956. Observations on growth rates of human tumors. Am J Roentgenol Radium Ther Nucl Med, 76:988–1000.

26.

Friberg

, Mattson

. 1997. On the growth rates of human malignant tumors: implications for medical decision making. J Surg Oncol, 65:284–297.

27.

Tuttle

, Brose

, Grande

, Kim

, Tahara

, Sabra

. 2017. Novel concepts for initiating multitargeted kinase inhibitors in radioactive iodine refractory differentiated thyroid cancer. Best Pract Res Clin Endocrinol Metab, 31:295–305.

28.

Sabra

, Sherman

, Tuttle

. 2019. Prolongation of tumour volume doubling time (midDT) is associated with improvement in disease-specific survival in patients with rapidly progressive radioactive iodine refractory differentiated thyroid cancer selected for molecular targeted therapy. Clin Endocrinol (Oxf), 90:617–622.