A 47-year-old man underwent a total thyroidectomy, central and right lateral neck dissection for a 5 cm right thyroid nodule in May 2002. Pathology revealed papillary thyroid carcinoma of the right nodule with vascular invasion and positive lymph nodes in the right central compartment (10/16 lymph nodes positive) and right lateral compartment (9/19 lymph nodes positive). In September 2002, he received radioactive iodine ablation with 146.7 mCI of I-131 after surgery with post-therapy I-131 scan showing thyroid bed uptake as well as uptake suspicious for residual lymph node metastases in the right lateral cervical lymph node region. On 6-month follow-up in July 2003, his stimulated thyrolobulin (Tg) levels were elevated at 251 ng/ml with negative Tg antibodies. A pretherapy scan did not show any uptake. He was subsequently treated 200 mCI of I-131 in August 2003 and post-therapy scan showed two foci of uptake in right lower neck consistent with residual lymph node metastases. Subsequent to this treatment, his stimulated Tg level was elevated to 4,734 ng/ml with negative Tg antibodies. Additional tests at this time including MRI neck, CT scan of head, neck, abdomen, and pelvis and FDG-PET showed FDG-positive lymph nodes in the right lateral neck and a 3 cm mass in the left thyroid bed. A bone scan done at this time was negative. He underwent a modified radical neck dissection in which 8 out of 21 lymph nodes were positive for papillary thyroid carcinoma. After surgical debulking, he was treated with 431.5 mCI of I-131 following a radioiodine dosimetry protocol. His post-therapy scan showed abnormal tracer uptake in left lateral neck region without anatomic correlate on cross-sectional imaging. He continued to have elevated Tg levels on thyroid hormone suppression with no obvious source on various imaging studies. In June 2007, PET-CT showed a new nasopharyngeal mass and a 3 mm right apex lung nodule. A stimulated Tg at this time was 661 ng/ml. He received radiation therapy (surgical debulking was not possible) to this nasopharyngeal mass without an improvement in Tg levels. His PET/CT in 2008 showed new sclerotic lesions in the right superior pubic ramus and body of the sternum. His Tg was 80 ng/ml on levothyroxine suppression, stimulated Tg was 1,138 ng/ml and a diagnostic I-131 scan was negative. He was followed conservatively for a few months at which time suppressed Tg levels continued to rise and his lung nodule grew to more than 1 cm in size. He was started on an experimental treatment protocol with sunitinib in May 2009. He completed 11 monthly cycles of sunitinib in June 2010 at which time his Tg had risen to 1,039 ng/ml and PET/CT showed stable pelvic and sternal lesions but a new occipital bone lesion and new bilateral subpleural lung metastases had developed. He was withdrawn from the sunitinib protocol for progressive disease and started on monthly zoledronic acid infusions for his bone metastases. The occipital bone lesion was surgically resected and treated with external beam radiotherapy. A repeat PET/CT in August 2011 showed an increase in size of his pelvic and sternal lytic lesions. He received radiotherapy to these lesions and enrolled in phase-2 study with lenvatinib. Zoledronic acid was discontinued and he began therapy with monthly subcutaneous denosumab for his bone metastases. His last imaging showed stable lung and bone disease.

Bone metastases are uncommon in thyroid cancer. While follicular thyroid cancer accounts for less than 15 % of all differentiated thyroid cancers, bone metastases occur in up to 7–20 % of these patients. Bone metastases are less common in papillary thyroid cancer, but still may be seen in up to 1–7 % of cases. In absolute terms however, the number of patients with bone metastases due to papillary thyroid cancer is higher since papillary thyroid cancer is more common. Overall, the incidence of bone metastases in well-differentiated thyroid cancer is 2–13 %. Skeletal metastases often are clinically silent but can present with pain, pathologic fracture, painful radiculopathy, bladder and bowel incontinence, and weakness of one more extremities from spinal cord compression. Spine is the most common site of bone metastases and 25 % of patients may have isolated bone metastases, with 15 % having both lung and bone metastases. It is important to be vigilant for metastatic disease since the presence of metastases lowers the 10-year survival rates for differentiated thyroid cancer from about 90 % to 40 %.

It can be difficult to detect bone metastases from thyroid cancer. Diagnostic I-131 scans are insensitive, while post-therapy scans perform better. X-rays and bone scintigraphy can be used in the evaluation of osseous metastases, but these imaging modalities are often limited by their poor specificity and their ability to detect disease only when more than half of an involved bone has been destroyed. Tc-99 bone scintigraphy can detect skeletal metastases earlier than plain radiographs when there is a predominant osteoblastic component to the lesion. But because thyroid cancer metastases are predominantly osteolytic, bone scintigraphy is of limited value in thyroid cancer with high false positive and false negative rates.

MRI of the whole body or of specific bones is an excellent modality to visualize the medullary component of bones and detailing the extra-skeletal extent of disease. Whole body MRI has higher sensitivity than PET/CT for diagnosis of osseous metastases (85–95 % vs. 62–91 %). CT alone is valuable in imaging cortical bone. In a prospective study of 80 patients comparing FDG PET/CT, I-131 SPECT/CT, and 99m Tc-MDP bone scans, FDG PET/CT and I-131 SPECT/CT were significantly superior to bone scans in detecting osseous metastases. PET scans have a role in predicting prognosis as well; a positive PET/CT increases the risk of death from thyroid cancer by fourfold.

Two future modalities for detecting bone metastases are 18F-fluoride PET/CT and Iodine-124. 18F-fluoride is a bone-seeking, positron emitting molecule with excellent sensitivity and specificity for detecting bone lesions, especially osteolytic metastases. When combined with PET/CT, 18F-fluoride can provide exquisite spatial resolution of osseous metastases with some studies suggesting better performance than PET/CT and Tc-99 bone scintigraphy. I-124 emits a positron that can be detected by PET scan. One study has demonstrated the superiority of I-124 PET/CT compared to I-131, I-124, and CT scans alone in detecting bone metastases.

We recommend that when extracervical spread is suspected, one should obtain an FDG-PET/CT. If bone metastases are discovered, then a directed MRI or CT should be employed to specifically define the lesion(s) of interest and as an aid in planning any surgical approaches or use of other modalities in the treatment of destructive osseous lesions.

In most situations where the histology of primary malignancy has been established, it is usually unnecessary to biopsy new lesions that present at distant sites. However, if a bone lesion represents the first manifestation of recurrent thyroid cancer, biopsy is recommended. Bone biopsy is not necessary if the bone lesions take up radioactive iodine on diagnostic or post-therapy I-131 scans or if a patient with widespread disease has undergone bone biopsy before confirming metastases of thyroid origin.

A needle biopsy is recommended for newly detected spine and pelvis bone lesions that do not take up radioiodine. Sections of biopsy should be carefully examined for histological subtype (papillary, follicular, Hürthle cell, and poorly differentiated carcinomas). Special stains for thyroid transcription factor (TTF), Tg, cytokeratin, and calcitonin should be employed and the need for supplemental stains like prostate-specific antigen should be individualized. Another simple technique that can be used in evaluating distant metastases is detection of thyroglobulin in the washout of fine needle aspirates of nonthyroidal masses. While not specifically studied in bone metastases, the value of the technique is probably similar. Once a bone biopsy has confirmed the origin of the tumor, I do not recommend biopsies of subsequent skeletal lesions except in rare circumstances.

There is no prospective study evaluating the effect of I-131 therapy on survival in patients with metastatic differentiated thyroid cancer. A retrospective study of 444 patients by Durante et al. showed that only 7 % of patients with bone metastases achieved remission with I-131 therapy and even 43 % of patients who had iodine avid lesions still did not achieve remission with radioactive iodine therapy. The factors that predicted response to I-131 therapy were age <40 years, solitary lesion and well-differentiated cancer. 10-year survival rates for complete responders, partial responders, and nonresponders were 92 %, 29 %, and 10 %, respectively.