Patients

The key eligibility criteria are as follows: age >18 years, Eastern Cooperative Oncology Group performance status (ECOG PS) score 0–2, measurable pathologically confirmed differentiated thyroid carcinoma, lesion refractory to radioiodine treatment and no prior therapy with tyrosine kinase inhibitors including sorafenib and lenvatinib. The definition of refractory to radioiodine treatment is described below. All patients will have received thyroid hormone replacement therapy with thyroid-stimulating hormone suppression.

The exclusion criteria are as follows: medullary or anaplastic carcinoma, prior treatments (including molecular targeted agents and/or conventional chemotherapy), other synchronous malignancies other than early stage lesions curable via endoscopy, fasting blood glucose ≥150 mg/dL; ECOG PS 3–4; thromboembolism requiring treatment, pregnant or nursing women, and any other reasons as judged by the investigators.

Study design

This will be a non-randomised multicentre study. The study protocol has been approved by all relevant institutional review boards; all recruited patients are required to provide written informed consent. This trial is registered in the University Hospital Medical Information Network (UMIN) Clinical Trials Registry (number 000022592), and is being conducted in accordance with the Declaration of Helsinki. Patient recruitment commenced on 1 June 2016 and will continue through 31 March 2019. The data are estimated to be available no earlier than April 2020 considering the median PFS of patients receiving lenvatinib.

Eligible patients will undergo FDG PET/CT, contrast-enhanced CT, blood tests for free tri-iodothyronine (fT3), free thyroxine (fT4), TSH, thyroglobulin (TG) and other baseline values within 28 days prior to the initiation of treatment. Next, eligible patients will receive treatment with oral lenvatinib 24 mg once daily. One week after treatment initiation, patients will undergo FDG PET/CT and repeat blood testing. Contrast-enhanced CT will be performed at least 4 weeks after the initiation of treatment as the gold standard for evaluation. Treatment response will further be confirmed by additional CT performed at least 8 weeks thereafter. Patients will be observed until disease progression or death (see figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Flowchart of the study. ¹⁸F-FDG PET/CT, ¹⁸F-fluorodeoxyglucose positron emission tomography/CT; T3, tri-iodothyronine; T4, thyroxine; TSH, thyroid-stimulating hormone.

Refractoriness to radioiodine treatment

In this study, a patient is considered refractory to radioiodine treatment if any of the following apply:

• At least one measurable lesion without iodine uptake is visible on any radioiodine scan.

• At least one measurable lesion has progressed after radioiodine therapy despite radioiodine avidity at the time of treatment.

• Progression persists despite a cumulative activity of radioiodine that is >400 mCi (14.8 GBq).

Accumulation of administered ¹³¹I in the metastatic lesion may not be sufficient because most of the radioactive iodine accumulates in the thyroid bed on first treatment. Hence, patients with metastatic lesions receive radioiodine treatment at least twice; if there is no uptake in the lesion on radioiodine scanning after two treatments, the patient can be judged to be refractory to radioiodine treatment. The 2015 American Thyroid Association Management guideline does not mention cumulative radioiodine activity.9

Primary objective

The highest metabolic activity within the tumour (ie, the maximum standard uptake value (SUVmax)) obtained by PET/CT is considered the most common indicator of metabolic change in a tumour. The primary goal is to evaluate the discriminative power of the change of SUVmax, which is obtained via PET/CT 1 week after the initiation of lenvatinib treatment, and compare it to the response evaluation obtained by contrast-enhanced CT performed at least 4 weeks after the initiation of treatment with lenvatinib.

Secondary objective

One of the secondary objectives is to determine the optimal cut-off value for the change of SUVmax, which is determined by calculating the shortest distance from the upper left corner of the plot to the ROC curve. The optimal cut-off value for SUVmax obtained from baseline PET/CT is also determined, and the discriminative power of the SUVmax obtained from baseline PET/CT is examined. Another secondary objective is to perform the analysis above by using total lesion glycolysis (TLG), which is the product of the mean SUV and metabolic tumour volume (MTV). The MTV is the sum of the tumour volumes that show higher SUVs than a predetermined threshold. In this study, we will use various thresholds (SUV ≥2.5, ≥3.0, ≥3.5 and ≥4.0) because that which best delineates the tumour remains unclear.10 11 Additionally, parameters obtained by texture analysis of PET/CT, reflecting the underlying spatial variation and heterogeneity of voxel intensities within a tumour, will be analysed to assess their ability to yield predictive and prognostic values. The response rate, PFS and dose intensity of lenvatinib will also be investigated.

Patients will not be excluded from the analysis if the dose of lenvatinib is reduced between the two FDG PET/CT scans. The rates of change in fT3, fT4, TSH and TG will also be calculated and analysed in terms of treatment outcomes.

Image acquisition and evaluation

Although this study is being conducted as a multicentre study, all PET/CT scanning will be performed using the same device at Hokkaido University Hospital for consistency. FDG is produced using an in-house cyclotron or will otherwise be purchased from Nihon Medi-Physics Co, Ltd (Tokyo, Japan). Patients are instructed to fast for at least 6 hour before ¹⁸F-FDG administration to achieve a blood glucose level <150 mg/dL. The patients’ blood glucose levels are measured before ¹⁸F-FDG administration. ¹⁸F-FDG (3.7 MBq/kg) is administered intravenously, and patients then rest in a quiet and dim room. Approximately 50 min after ¹⁸F-FDG injection, an integrated PET/CT system (Gemini TF64; Hitachi Medical Corporation Ltd, Tokyo, Japan) begins to acquire images. PET/CT is performed in accordance with guidelines published by the National Cancer Institute.12 PET/CT scanning is typically performed from the level of the vertex of the skull or orbits to the upper thighs unless the patient has known metastatic lesions (eg, tibia lesions). A whole-body unenhanced CT scan is performed before emission acquisition for the purposes of anatomic localisation and attenuation correction. CT parameters include an axial slice thickness of 3.75 mm, a tube voltage of 140 kV, fixed (120 mA) or variable tube current, and table speed of 13.5 mm/s. Emission data are acquired in a three-dimensional mode with a state-of-art time-of-flight technique and are stored in list mode. The emission scanning is initiated 60 min (±5 min) after FDG administration for 3 min per bed position. The transaxial and axial fields of view are 576 mm and 180 mm, respectively. PET images are reconstructed for attenuation-corrected and non-attenuation-corrected data using an ordered-subset expectation maximisation algorithm. PET, CT and fused images are displayed in 5 mm slices. For contrast-enhanced CT, examinations are performed at each institution. The reconstructed image data are stored electrically in the hospital information system at Hokkaido University Hospital, while the list mode data are stored electronically in the Department of Radiology at Hokkaido University Hospital.

Image review and tumour analysis

FDG PET/CT and contrast-enhanced CT data are interpreted clinically by a board-certified nuclear medicine physician and a board-certified diagnostic radiologist, respectively (both of whom are board-certified), at Hokkaido University Hospital. For FDG PET/CT analysis, a non-physiological increase in FDG uptake compared with the surrounding background is classified as positive for disease. The nuclear medicine physician measures SUVmax and SUVmean within the tumour using a spherical volume-of-interest. The SUV is defined as measured activity concentration (kilobecquerels per millilitre) multiplied by body weight (grams) divided by injected activity (kilobecquerels). Metabolic response is classified as a decrease in SUV of over 30%. Contrast-enhanced CT is analysed based on the Response Evaluation Criteria in Solid Tumours.

Sample size

As the response rate of lenvatinib was previously reported to be 64.8%,13 we assumed expected response/non-response rates of 60% and 40%, respectively, indicating a 6-to-4 ratio. Obuchowski’s method enables us to compute sample sizes for one and two ROC curves14 15; this method was used to calculate the required number of responses.14 15 The expected discriminative power of SUVmax is the area under the curve (AUC), which is 0.84, and its variance is defined using Obuchowski’s method where . Using this method, we calculated the number of required responding and non-responding subjects by assuming a two-sided significance level of 5%. Given that the diagnostic ability of the SUVmax is the AUC (0.84) and that the null hypothesis is AUC=0.5, 20 patients are ultimately required to achieve a power of 80%, including 12 and 8 in the response and non-response groups, respectively. Assuming 10% of the patients would be lost to follow-up or would discontinue treatment within 1 week of initiation of lenvatinib, we aim to include 22 patients. Sample size calculation was performed by the power.t.test package of R studio.

Statistical considerations

Statistical analyses are performed by using RStudio (V.0.99.485 or later, RStudio). ROC analyses will be performed to assess whether the change in the PET-CT-based SUVmax between 1 and 4 weeks after the initiation of treatment can discriminate response/non-response at 4 weeks. The AUC of the ROC curve is calculated using the trapezoidal method, and 95% CIs are also calculated. The same methods are used for calculating the AUCs of the ROC curves for secondary endpoints. The optimal cut-off is calculated using the Youden Index. Sensitivities, specificities and positive and negative predictive values will also be calculated along with the corresponding 95% CIs of the cut-off values. Similar analyses will be performed for the primary endpoint as well as for TLG. The PFS rates and their 95% CIs will be calculated using the Kaplan-Meier method and Greenwood formula, respectively.

Patient and public involvement

The research question was developed in response to discussions about whether FDG PET/CT can predict the treatment outcome and avoid adverse effects and high treatment costs in advance. No specific patient advisers or organisations were involved in the design or conduct of the study. The study results will be disseminated to study participants in person or by post based on their preferences.