• thyroid carcinoma;
  • histopathologic characterization;
  • radioactive iodine;
  • fluorodeoxyglucose-positron emission tomography (FDG-PET)


  1. Top of page
  2. Abstract


Radioactive iodine-refractory (RAIR) 18F-fluorodeoxyglucose (FDG)-positron emission tomography (PET) positive thyroid carcinomas represent the major cause of deaths from thyroid carcinomas (TC) and are therefore the main focus of novel target therapies. However, to the authors' knowledge, the histology of FDG-PET-positive RAIR metastatic thyroid carcinoma has not been described to date.


Metastatic tissue from RAIR PET-positive patients identified between 1996 and 2003 at the study institution were selected for histologic examination. The biopsied metastatic site corresponded to a FDG-PET positive lesion sampled within 2 years (87% of which were sampled within 1 year) of the PET scan. Detailed microscopic examination was performed on the metastatic deposit and the available primary tumors. Poorly differentiated thyroid carcinomas (PDTC) were defined on the basis of high mitotic activity (≥5 mitoses/10 high-power fields) and/or tumor necrosis. Other types of carcinomas were defined by conventional criteria. The histology of the metastases and primary were analyzed, with disease-specific survival (DSS) as the endpoint.


A total of 70 patients satisfied the selection criteria, 43 of whom had primary tumors available for review. Histologic characterization of the metastasis/recurrence in 70 patients revealed that 47.1% (n = 33 patients) had PDTC, 20% (n = 14 patients) had the tall cell variant (TCV) of papillary thyroid carcinoma, 22.9% (n = 16 patients) had well-differentiated papillary thyroid carcinoma (WDPTC), 8.6% (n = 6 patients) had Hurthle cell carcinoma (HCC), and 1.4% (n = 1 patient) had anaplastic carcinomas. The histopathologic distribution of the tumor in the primaries was: PDTC, 51%; TCV, 19%; WDPTC, 23%; and widely invasive HCC, 7%. A differing histology between the primary tumor and metastasis was observed in 37% of cases (n = 16 patients). In the majority of instances (63%; 10 of 16 patients) this was noted as transformation to a higher grade. Of the primary tumors classified as PTC, 70% progressed to more aggressive histotypes in the metastasis. Tumor necrosis and extensive extrathyroid extension in the primary tumor were found to be independent predictors of poorer DSS in this group of patients (P = .015). Approximately 68% of the PDTC primary tumors were initially classified by the primary pathologist as better-differentiated tumors on the basis of the presence of papillary and/or follicular architecture or the presence of typical PTC nuclear features.


Several observations can be made based on the results of the current study. The majority of metastases in patients with RAIR PET-positive metastases are of a histologically aggressive subtype. However, well‒differentiated RAIR metastatic disease is observable. Poorly differentiated disease is underrecognized in many cases if defined by architectural and nuclear features alone. The presence of tumor necrosis was found to be a strong predictor of aggressive behavior, even within this group of clinically aggressive tumors. Finally, there is a significant amount of histologic plasticity between primary tumors and metastases that may reflect the genetic instability of these tumors. Cancer 2008. © 2008 American Cancer Society.

Thyroid carcinomas of follicular origin demonstrate great morphologic diversity, ranging from well-differentiated tumors with an indolent course to undifferentiated (anaplastic) carcinomas with a dismal prognosis. Well-differentiated tumors, characterized by a papillary and/or follicular architecture on histologic examination, usually have an excellent prognosis. At the opposite pole, undifferentiated thyroid carcinoma provides few clues as to its histogenesis and resembles high-grade sarcoma or poorly differentiated squamous cell carcinoma. Nearly all patients with these undifferentiated thyroid carcinomas die of progressive disease within a matter of weeks to months from the time of diagnosis.1 Between these 2 extremes exists a group with an intermediate prognosis called poorly differentiated thyroid carcinoma (PDTC). Defining criteria for the diagnosis of PDTC is a source of controversy.2 Some authors have advocated using architectural features such as a solid/trabecular growth pattern (as advocated by Japanese pathologists), whereas others propose using a proliferate grading system (ie, nuclear atypia, mitoses, and necrosis) to establish the diagnosis of PDTC.3–6 Recently, an international group of pathologists convened in Turin, Italy and proposed new consensus diagnostic criteria for PDTC. These authors require the presence of 1) a solid/trabecular/insular growth pattern; 2) the absence of the nuclear features of papillary thyroid carcinoma (PTC); and 3) the presence of at least 1 of the following features: convoluted nuclei, high mitotic activity, and tumor necrosis. We share with the Turin proposal the need to include mitosis and tumor necrosis in the definition of PDTC because in our opinion and theirs these parameters are strong negative prognostic indicators. However, we believe these high-grade features by themselves are sufficient to label a tumor as PDTC (irrespective of growth pattern and nuclear morphology), whereas the Turin proposal requires a solid growth pattern and the absence of the nuclear features of PTC. Both our group and the Turin panel include ‘insular carcinoma’ in their definition because the vast majority of insular carcinoma fits both definitions.7

The majority of patients with nonanaplastic thyroid cancer will have disease-free survival rates of >30 years with appropriate surgery and radioactive iodine (RAI) therapy. However, 2% to 5% of these tumors will lose their differentiated features, become difficult to monitor by serum thyroglobulin, develop recurrent non-RAI-avid 18F-fluorodeoxyglucose (FDG)- positron emission tomography (PET)-positive disease, and lead to the patient's death.8 This change in biologic behavior has been attributed to the presence of poorly differentiated elements in the metastatic tumor of these patients.9 However, to our knowledge, no study to date has been performed to examine the histology of FDG-PET‒positive, RAI-refractory (RAIR) metastatic thyroid carcinomas.

Instead, the majority of studies have attempted to correlate FDG-PET status at the time of disease recurrence with the histology of the primary surgical resection in thyroid carcinoma. This methodology is problematic from a biologic standpoint because the metastatic/recurrent disease may have undergone considerable genomic and histologic changes during the course of time. This has important therapeutic implications vis-a-vis targeted therapy because it is the metastatic/recurrent disease that is being treated. Ergo, in the current study, we characterized the histology of metastatic RAIR PET-positive thyroid carcinoma. In addition, we compared the histology of the metastases with the available primary resection specimens. These data help in understanding the progression of thyroid carcinoma and may have implications in the development of novel targeted therapies aimed at curing RAIR disease, which is the major cause of death in patients with thyroid carcinomas.


  1. Top of page
  2. Abstract

Patient Population and Inclusion Criteria

We previously reviewed all FDG-PET scans of patients with follicular derived thyroid carcinoma between January 1996 and June 2003 whose PET scans were available for independent review.10 The majority of patients referred for PET scanning had RAIR disease. A patient was deemed RAIR if the patient had an elevated serum thyroglobulin with structural disease in the setting of a negative radioiodine whole-body scan. In addition, FDG-PET scanning is routinely obtained in the initial staging and follow-up of Hurthle cell carcinoma (HCC).

A patient was entered in the study if: 1) he/she had RAIR, FDG-PET‒positive, non-RAI-avid metastatic/recurrent thyroid carcinoma; 2) the biopsy of the metastasis/recurrence corresponded to a lesion identified as PET-avid; and 3) the biopsy of the metastasis/recurrence was performed within 2 years of the FDG-PET and its histologic slides were available for re-review by 2 pathologists (R.G. and M.R.).

The study was reviewed and approved by the Memorial Sloan‒Kettering Cancer Center Institutional Review Board.


Images were acquired on a General Electric Advance PET camera, or dedicated inline PET/computed tomography (CT) scanners (Biograph; Siemens/CTI, Knoxville, Tenn; or Discovery LS; GE Medical Systems, Waukesha, Wis) and analyzed as previously described.10 Briefly, patients were fasted for at least 6 hours before intravenous injection of 10 to 15 millicurie (mCi) (370-555 megabecquerel [MBq]) of FDG. When patients were imaged with a combined PET/CT, the associated CT images were evaluated in the same planes for anatomic reference. The scans were inspected visually and the sites of abnormally increased radiotracer uptake were counted and grouped by location. In addition, the standardized uptake value of the lesion with the highest FDG uptake in the body was identified (SUVmax). The FDG-PET images were read on 2 separate occasions, once by a team of 2 experienced nuclear medicine physicians and later by an independent senior nuclear medicine physician. The image reviewer was blinded to clinical and laboratory information. The site of an individual metastatic lesion was grouped as neck (above sternal notch and below ears), mediastinum, lungs, bone, and other (eg, brain, liver, or adrenal). Whenever the site of a lesion could not be determined, it was assigned to the ‘other’ category.

Whole-body 131-I Scans

Whole-body radioiodine scans were conducted and interpreted as previously reported.11 The radioiodine scan had to be performed within 6 months of the FDG-PET scan. The most informative scan (either a diagnostic or post‒therapy scan) during the interval was used to determine radioiodine avidity status. All radioiodine scans were read by a nuclear medicine physician and the results were taken from their official reports.


Two board-certified pathologists with a special interest in thyroid carcinoma (R.G. and M.R.) examined primary and metastatic sites from RAIR PET-positive patients. The pathologists were blinded with regard to the survival status of the patients. The biopsied metastatic site corresponded to an FDG-PET positive lesion sampled within 2 years of the FDG-PET scan. The tumors were classified according to World Health Organization 2004 criteria with the exception of the tall cell variant (TCV) of PTC and PDTCs.12 PDTCs were defined by proliferative grading features, which we defined as ≥5 mitoses/10 high-power fields (HPF) and/or tumor necrosis regardless of architectural pattern.4 Tumors were classified as TCV if they contained ≥50% tall cells (ie, cells whose height is at least twice their width with ample eosinophilic cytoplasm and the characteristic nuclear features of PTC). In addition to histologic typing, the following histopathologic parameters were recorded in the primary tumor. Tumor size was measured as the maximum dimension of the resected tumor specimen. The mitotic rate was determined by counting 10 HPF (×400) using an Olympus microscope (U-DO model; Olympus America, Melville, NY) in the areas with the greatest concentrations of mitotic figures. Fresh tumor necrosis was classified as absent, focal (≤5% of the tumor area), or extensive (>5% of the tumor area). The presence or absence of a tumor capsule was recorded. Vascular and capsular invasion were identified according to the criteria outlined in the most recent Armed Forces Institute of Pathology (AFIP) fascicle regarding thyroid tumors.1 Vascular and capsular invasion were subdivided into 3 categories: absent, focal (<4 invasive foci), and extensive (≥4 invasive foci). The type of tumor cells present in the PD carcinoma was classified as papillary-like (if the nuclear features resembled those of PTC), follicular-like (if the cell resembled a follicular carcinoma cell), or oncocytic (if the cell had the nuclear features and granular eosinophilic cytoplasm of so-called Hurthle cells). In PD carcinoma the growth pattern (follicular, papillary, solid/trabecular) and its extent were recorded.

The presence or absence of extrathyroid extension of the tumor into the stroma of the perithyroid tissues was recorded as well as the presence or absence of vascular invasion in extrathyroid vessels. Finally, microscopic resection margins were categorized as positive (tumor at the inked margin) or negative (no tumor at the inked margin). Regional lymph node status was also recorded.

Laboratory Data

Serum thyroid‒stimulating hormone and thyroglobulin were measured as previously reported.11

Statistical Analysis

Disease-specific survival, as measured from the time of diagnosis to date of death or last clinical follow-up, was the primary endpoint of this study. Disease-specific survival from the time of FDG-PET scan was also recorded and analyzed. Survival probabilities were estimated by the Kaplan-Meier method and compared using the log-rank test. Multivariate analysis was performed using the Cox regression model. All statistics were performed using GraphPad (GraphPad Prism 5.01; GraphPad, San Diego, Calif) and STATA software (StataCorp, College Station, Tex).


  1. Top of page
  2. Abstract

Histology of Metastases

Histologic characterization of metastases/recurrence in 70 RAIR PET-positive thyroid carcinoma patients revealed that 47.1% had PDTC, 20% had TCV of PTC, 22.9% had well-differentiated PTC (including classic and follicular variants), 8.6% had HCC, and 1.4% had anaplastic carcinomas (Table 1). We histologically examined 76 RAIR PET-positive metastatic/recurrent lesions in these 70 patients. Of these 76 lesions, 30 (39%) were neck lymph node metastases, 18 (24%) were cervical soft tissue recurrences, and 28 (37%) were distant metastases, including 11 bone metastases (Table 2).

Table 1. Clinicopathologic Characteristics of 70 Cases of Metastatic Radioactive Iodine-Refractory, PET-Positive Thyroid Carcinoma
  1. PET indicates positron emission tomography; PD, poorly differentiated; PTC, papillary thyroid carcinoma; TCV, tall cell variant of papillary thyroid carcinoma; HCC, Hurthle cell carcinoma; ANA, anaplastic carcinoma; DOD, died of disease; AWD, alive with disease; AND, alive with no evidence of disease.

No. of cases33 (47.1%)16 (22.9%)14 (20.0%)6 (8.6%)1 (1.4%)
Median age at diagnosis, y6246.55342.553
Gender ratio, M:F15:188:86:86:01:0
No. DOD24 (73%)3 (19%)3 (21%)2 (33%)1 (100%)
No. AWD9 (27%)8 (50%)7 (50%)4 (67%)0
No. AND05 (31%)4 (29%)00
Table 2. Distribution of Metastatic Sites in 70 Cases of Metastatic Radioactive Iodine-Refractory, PET-Positive Thyroid Carcinoma
 Neck LN (n=30)Neck ST (n=18)Mediastinum (n=5)Lung (n=7)Chest wall (n=2)Liver (n=1)Bone (n=11)Brain (n=1)Extremity ST (n=1)
  1. PET indicates positron emission tomography; LN, lymph node; ST, soft tissue; WD, well differentiated; TCV, tall cell variant of papillary thyroid carcinoma; HCC, Hurthle cell carcinoma; PD, poorly differentiated; ANA, anaplastic carcinoma.

No. with WD731100401
No. with TCV830300000
No. with HCC200010100
No. with PD12114311610
No. with ANA110000000

The cohort consisted of 33 females and 37 males with a mean age at diagnosis of 53 years (range, 22–85 years). At the time of last follow-up, 33 (47%) patients were dead of disease, 28 (40%) were alive with disease, and 9 (13%) were alive with no evidence of disease. The median survival from the time of primary tumor resection in patients who died of disease was 6.5 years. The median follow-up from the time of primary resection was 8.9 years.

In all cases, the biopsied site corresponded to a lesion identified on a PET scan performed within 2 years of the biopsy. It is interesting to note that 61 of the 70 patients (87%) were sampled within 1 year of the FDG-PET scan. The mean and median SUV of the sampled metastatic/recurrent sites were 7.6 and 6.3, respectively. In addition, the sampled metastasis corresponded to the lesion with the highest SUV (SUVmax) on the PET scan in 36% of cases (n = 25 patients).

Histology of Paired Primary and Metastasis

Slides of both the primary and metastasis/recurrence were available for examination in 43 cases (Table 3). A differing histology between the primary and metastasis was observed in 37% (n = 16) of cases. In most instances (63%; 10 of 16 patients), this was noted as transformation to a higher grade (eg, transformation to PDTC from a follicular variant PTC. (Fig. 1). Less often (38%; 6 of 16 of patients), a better-differentiated, low-grade metastasis was observed in which the primary tumor was PDTC or TCV.

thumbnail image

Figure 1. Microscopic pictures of hematoxylin and eosin (H&E)-stained sections of an infiltrative follicular variant (FV) of papillary thyroid carcinoma (PTC) that was well differentiated (A-C), occurring in a 55-year-old woman whose disease progressed to poorly differentiated thyroid carcinoma (PDTC) in a skull metastases 15 years after diagnosis (D and E: microscopic view; F: positron emission tomography scan [PET]). (A) Follicle formation at low power in FVPTC. (B) Nuclear clearing and overlapping typical of PTC (arrow). (C) Extrathyroid vascular invasion in FVPTC (arrow indicates involved, medium-size vessel). (D) Trabecular growth pattern in the PDTC (arrow). (E) Necrosis (n) in PDTC. (F) PET-positive skull metastases (arrow). The patient died of thyroid carcinoma 1 year and 1 month after her PET scan.

Download figure to PowerPoint

Table 3. Paired Histologic Data of Primary Tumors and Metastases
Primary tumor type (n=43)Histotype of metastasis, no.Histotype differences between primary tumor and metastases, no.Histopathologic disease progression, no.
  1. PDTC indicates poorly differentiated carcinoma; PD, poorly differentiated; TCV, tall cell variant of papillary thyroid carcinoma; PTC, classic papillary thyroid carcinoma and follicular variant; HCC, Hurthle cell carcinoma; ANA, anaplastic carcinoma.

22 PDTC17 PD (77.3%)5 (22.7%)0
2 TCV (9.1%)
3 PTC (13.6%)
3 HCC2 HCC (66.7%)1 (33.3%)1 (33.3%)
1 PD (33.3%)
8 TCV5 TCV (62.5%)3 (37.5%)2 (25%)
1 ANA (12.5%)
1 PD (12.5%)
1 PTC (12.5%)
10 PTC3 PTC (30%)7 (70%)7 (70%)
4 TCV (40%)
3 PD (30%)

Histology of Primary Tumors

PDTC carcinoma was diagnosed in 51.2% of primary resections (n = 22 patients) performed in patients with a mean age at diagnosis of 54 years (range, 22–78 years). There were 14 women and 8 men, with a mean tumor size of 4.7 cm. In the primary PDTC category, extrathyroidal extension was identified in 50% of cases (n = 11 patients) and was extensive in 36.4% (n = 8 patients). Vascular invasion was observed in 68% of cases (n = 15 patients) and was found to be extensive in 50% of cases (n = 11 patients). In 1 instance, invasion of extrathyroidal vessels was present. Lymph node metastases were histologically documented in 50% of cases (n = 11 patients) at the time of presentation. In addition, 59.1% of patients (13 of 22 patients) diagnosed with PDTC in the primary tumor site died of disease compared with 14.3% of patients (3 of 21 patients) with better-differentiated tumors diagnosed in the primary tumor site. PDTC was initially classified by the primary pathologist as a better-differentiated tumor in 68.2% of cases (15 of 22 patients) (Fig. 2). Fourteen of these 15 cases (93.3%) had a significant (range, 10–100%) proportion of tumor comprised of follicular/papillary architecture and 9 cases (60%) demonstrated the nuclear features of classic PTC.

thumbnail image

Figure 2. Microscopic pictures of hematoxylin and eosin (H&E)-stained sections of a poorly differentiated thyroid carcinoma (PDTC) occurring in a 65-year-old woman that initially was classified as papillary thyroid carcinoma (PTC). (A) Low-power view showing a tumor predominantly growing in follicles with tumor necrosis (marked by an arrow). (B) High-power view of the same field demonstrating overlapping clear nuclei reminiscent of PTC (arrow) but with central necrosis (n). This patient died of thyroid carcinoma 12.5 years after diagnosis.

Download figure to PowerPoint

The TCV of PTC was identified in 18.6% of primary tumors (n = 8 patients). The TCV group was comprised of 3 women and 5 men with a mean age at diagnosis of 60 years (range, 49-85 years). The average tumor size was 3.6 cm. All patients had extrathyroid extension and it was found to be extensive in 87.5% (n = 7 patients). Vascular invasion was noted in only 1 instance (12.5%) and lymph node metastases was observed in 75.0% of patients at the time of presentation (n = 6 patients). Two cases progressed (1 to PDTC, 1 to anaplastic) and 1 case recurred as well differentiated PTC.

Well-differentiated PTC (encompassing classic PTC and the follicular variant of PTC) was identified in 23.3% of primary resections (n = 10 patients), with progression to a more aggressive histology in the metastasis noted in 70% of cases. Nearly all cases of primary classic PTC had either high-risk clinical features (ie, advanced age) or aggressive invasive histology features at the time of presentation (ie, extensive extrathyroid extension or extrathyroid vascular invasion) on pathologic examination (Table 4).

Table 4. Clinicopathologic Characteristics of Primary Well-Differentiated Papillary Thyroid Carcinoma (Classic Papillary and Follicular Variant)
Age, Years/SexPrimary diagnosisSize, cmETEVIMetastases histologyOutcome
  1. ETE indicates extrathyroid extension; VI, vascular invasion; F, female; cPTC, classic papillary carcinoma; TCV, tall cell variant of papillary thyroid carcinoma; AWD, alive with disease; PTC, well differentiated papillary thyroid carcinoma; NA, not available; M, male; PDTC, poorly differentiated papillary thyroid carcinoma; PFV, follicular variant of papillary carcinoma.

25/FcPTC1.4Present in fatNoneTCVAWD
52/FMulticentric cPTC0.5-1Extensive in fat and muscleNonePTCAWD
29/FcPTC with tall cell featuresNAFocal in fatNoneTCVAWD
53/FcPTC3Focal in fatNoneTCVAWD
53/FcPTCAt least 2.6 cmNoneNoneTCVAWD
32/FcPTC2Focal in fatExtrathyroid VIPTCAWD
42/FcPTC3Extensive in fatExtensivePTCAWD
66/McPTCNAExtensive in fatNonePDTC (at presentation)AWD
55/FPFV, infiltrative5Extensive in fatExtrathyroid VIPDTCAWD

In those cases with primary PDTC, there was a significantly higher proportion of distant metastases (15 of 27 patients; 55.6%) compared with better-differentiated tumors (4 of 22 patients; 18.2%), with a P value of .008. Distant metastatic sites included the mediastinum, chest wall, liver, bone, and brain (Table 5). Finally, all 3 primary HCCs displayed extensive angioinvasion.

Table 5. Location of Metastatic Sites According to Histology of the Primary Tumor
 Neck LNNeck STMediastinumLungChest wallLiverBoneBrain
  1. LN indicates lymph nodes; ST, soft tissue; WD, well differentiated; TCV, tall cell variant of papillary thyroid carcinoma; HCC, Hurthle cell carcinoma; PD, poorly differentiated.

No. WD63000010
No. with TCV43011000
No. with HCC11001000
No. PD66340161

Disease-specific Survival

Survival from the time of primary resection in the 43 patients with primary tumor available for histologic examination

Extensive extrathyroid extension and the presence of tumor necrosis in the primary tumor were found to correlate with a decreased disease-specific survival (P = .015 for each variable) (Fig. 3). Median survival for cases with necrosis was 7.5 years and 16.2 years for those without. There was a trend toward decreased disease-specific survival in patients aged >45 years at time of primary resection (P = .05). Stage of disease (I, II vs III/IV), sex, mitotic rate, and histologic subtype were not found to be predictive of survival in this group of patients. On multivariate analysis, extensive extrathyroid extension and tumor necrosis remained the only independent predictors of poor survival (P = .022 and P = .006, respectively).

thumbnail image

Figure 3. Disease-specific survival from the time of surgical resection of the primary tumor according to tumor necrosis in the 43 patients with primary tumors that were available for histologic examination. The presence of tumor necrosis was found to be significantly correlated with decreased survival time. The median survival for groups with necrosis was 7.5 years and without was 16.2 years.

Download figure to PowerPoint

Survival from time of PET in the 70 patients with metastatic/recurrent tumor available for histologic examination

Neither sex nor histologic subtype of metastasis/recurrence was found to be associated with disease-specific survival from the time of PET. However, older age at PET (>45 years) was found to be associated with a statistically significant decrease in disease-specific survival (P = .0008). The SUV of the sampled lesion did not appear to impact disease-specific survival (P = .3229).


  1. Top of page
  2. Abstract

In the current study, we described the histology of RAIR PET-positive metastatic thyroid carcinoma and the primary tumors from which they derived. The largest group responsible for RAIR PET-positive entities was PDTC. The majority of PDTC patients presented with stage III or IV disease, as expected for this aggressive group of tumors.4 It is interesting to note that only 32% of these cases were initially diagnosed as PDTC in the primary tumor. Indeed, 9 cases had PTC nuclear features and 14 cases demonstrated a significant proportion of tumor that was comprised of a papillary/follicular architecture. These were initially diagnosed as classic PTC, follicular variant and moderately differentiated PTC despite the presence of tumor necrosis and/or high mitotic rate. This underscores the notion that in addition to architectural grading it is necessary to take into account ‘proliferative’ grading features (ie, necrosis and mitosis). Indeed, in this study, the presence of necrosis in the primary tumor was found to be an independent predictor of disease-specific survival, signifying that necrosis is a reliable marker of poorly differentiated behavior. A mitoses count of >5 per 10 HPF was not found to be able to significantly predict disease-specific survival in this group of patients with clinically aggressive disease, which may reflect a weaker prognostic value for this parameter. However, previous studies have shown a correlation between increased mitotic rate and the diagnosis of PDTC.4, 13 It is our opinion that identifying papillary/follicular architecture and/or nuclear PTC features does not exclude the possibility of making the diagnosis of PDTC and that the evaluation of mitosis and necrosis more accurately predicts the behavior of PDTC. PDTC diagnosed as papillary carcinoma (classic, follicular variant, or moderately differentiated) on the basis of growth pattern and nuclear features can mislead the clinician into believing she/he is dealing with a relatively slow-growing tumor that most likely will respond to RAI therapy. A classification of these tumors as PDTC may lead to a more intense follow-up (ie, PET scan).

Surprisingly, well-differentiated PTC was identified in 22.9% and 23.3% of metastatic and primary tumors, respectively. As shown in Table 4, 40% of the primary tumors demonstrated extensive extrathyroid extension. In addition, extrathyroidal vascular invasion was noted in 2 cases, highlighting the aggressive behavior of these tumors. The majority of well-differentiated PTC cases progressed to TCV or PDTC in the metastasis, accounting for the increased metabolic activity discovered on PET. It is interesting to note that 3 of the patients in this group were aged <30 years. In the majority of cases, well-differentiated carcinomas respond well to RAI, and overall younger patients have better long-term outcomes compared with older patients when matched for tumor size.14 Unfortunately, there appears to be no way to predict the behavior in these rare, well-differentiated RAIR tumors by morphology because extrathyroidal extension is not considered to be predictive of a poor response to RAI in young people. In such cases, the tumors may have an inability to concentrate iodine or may have a genetic profile that is similar to PDTC without demonstrating the classic phenotype (ie, they belie their true nature and appear well-differentiated, yet having a genetic profile akin to PDTC). This deserves further study at the molecular level.

The TCV of PTC was observed in 20.0% of metastasis and 18.6% of primary tumors. Although the TCV is known to present most often at a higher stage and cause a significant decrease in disease-free survival when compared with classic PTC, tumors confined to the thyroid do exist. These latter TCV without extrathyroid extension have a slightly increased rate of disease recurrence when compared with classic PTC without extrathyroid extension.15 However, in contrast to the aforementioned study, the cases of TCV in the current study demonstrated extensive extrathyroidal extension into adjacent fibroadipose tissue and/or skeletal muscle. One case progressed to anaplastic carcinoma and 1 case progressed to PDTC in the metastasis. It is interesting to note that progression to anaplastic thyroid carcinoma from a better-differentiated carcinoma was observed with TCV and not PD, as one might expect. An unusual variant of anaplastic thyroid carcinoma termed spindle cell squamous carcinoma had been previously suggested to occur in association with TCV.16 Whether progression to anaplastic thyroid carcinoma from TCV occurs more frequently than from other well-differentiated thyroid carcinoma in general is a matter for future investigation.

In the current study, 37% of patients demonstrated a significant degree of morphologic discordance between the primary tumor and the metastases. The majority of these differences are due to the fact that the tumor progressed at the metastatic site. This difference in morphology is most likely reflected at the genetic level, at which the number of chromosomal abnormalities increases as thyroid carcinomas progress. Indeed, our group showed that the median number of chromosomal abnormalities per case rises from 1 in well-differentiated PTC to 10 in anaplastic carcinoma, with intermediate values of 3 and 5 for TCV and PDTC, respectively.17 In a minority of cases, morphologic instability was best demonstrated in cases in which the metastasis demonstrated a better-differentiated histology compared with the primary tumor. This phenomenon has been previously reported in bone metastases from thyroid carcinomas.18 This observation might be explained by the limitations of performing biopsies on a limited number of metastatic sites, such that the site with PDTC was not sampled. However, in the current study, because all biopsied sites were PET-positive and RAI-negative, this appears unlikely. This unexpected finding could be because a PDTC genetic profile is being masked by a well-differentiated histology. As suggested by Heng,19 it may be that tumors are so genetically unstable that they may regain their lost genes as they progress. In a breast cancer cell line, Worsham et al20 found that several homozygous losses in key genes during the early stages were restored to their diploid status when the cell line became more virulent. Although the morphologic instability of these tumors is most likely paralleled by genetic heterogeneity between the primary tumor and metastases, the presence of mutations and genetic aberrations that are conserved along the thyroid carcinoma progression spectrum offers hope that these tumors may be amenable to target therapy. Indeed, B-RAF mutations were found in anaplastic carcinoma and the corresponding better-differentiated component, and we found that some chromosomal loss and gains are conserved in well‒differentiated PTC, PDTC, and anaplastic carcinoma.17, 21 Therefore, extensive molecular characterization of the metastases in a given patient taken together with morphologic and clinical findings would be the best approach to selecting which patients will respond best to targeted therapy.

In summary, although the majority of primary thyroid carcinomas leading to RAIR PET-positive metastatic disease are PDTC, well-differentiated tumors can also be responsible for RAIR disease. The majority of well-differentiated carcinomas will progress in the metastasis. A significant number of poorly differentiated, RAIR, PET‒positive tumors are underdiagnosed as well or moderately differentiated tumors because of the sole use of architectural grading. This emphasizes the need for careful grading of the primary tumor on the basis of proliferative features (ie, mitosis and necrosis) to follow patients more carefully. The importance of necrosis in the primary is reinforced by the finding that it was an independent predictor of poor disease-specific survival in this cohort of patients. RAIR PET-positive tumors display a significant amount of morphologic instability as they metastasize. This should lead the clinician to enter the patient into a clinical trial based on a biopsy of the metastatic site. Whether this morphologic instability is paralleled by genetic heterogeneity needs to be addressed by further studies aimed at the molecular profiling of these tumors. These genetic studies will have a major impact on the use of targeted therapies in RAIR PET-positive thyroid carcinomas.


  1. Top of page
  2. Abstract