Fluorodeoxyglucose Positron Emission Tomographic Scanning in ...

0021-972X/00/$03.00/0 The Journal of Clinical Endocrinology & Metabolism Copyright © 2000 by The Endocrine Society

Vol. 85, No. 3 Printed in U.S.A.

Prognostic Value of [18F]Fluorodeoxyglucose Positron Emission Tomographic Scanning in Patients with Thyroid Cancer* WEIPING WANG, STEVEN M. LARSON, MELISSA FAZZARI, SATISH K. TICKOO, KATHERINE KOLBERT, GEORGE SGOUROS, HENRY YEUNG, HOMER MACAPINLAC, JUAN ROSAI, AND RICHARD J. ROBBINS Nuclear Medicine and Endocrinology Services, Departments of Radiology (W.W., S.M.L., H.Y., H.M.), Medicine (R.R.), Medical Physics (K.K., G.S.), Epidemiology and Biostatistics (M.F.), and Pathology (S.T., J.R.), The Laurent and Alberta Gerschel PET Center, Memorial Sloan Kettering Cancer Center, New York, New York 10021 ABSTRACT Poorly differentiated thyroid cancer lesions often lose the ability to concentrate radioactive [131I]iodine (RAI) and exhibit increased metabolic activity, as evidenced by enhanced glucose uptake. We incorporated [18F]fluorodeoxyglucose (FDG) positron emission tomography (PET) scanning into the routine follow-up of a cohort of thyroid cancer patients undergoing annual evaluations. One hundred and twenty-five patients who had previous thyroidectomies were included. They had diagnostic RAI whole body scans, serum thyroglobulin measurements, and additional imaging studies as clinically indicated. During 41 months of follow-up, 14 patients died. Univariate analysis demonstrated that survival was reduced in those with age over 45 yr, distant metastases, PET positivity, high rates of FDG uptake, and high volume of the FDG-avid disease (⬎125 mL). Survival did not correlate with gender, RAI uptake, initial histology, or

C

grade. Multivariate analysis demonstrated that the single strongest predictor of survival was the volume of FDG-avid disease. The 3-yr survival probability of patients with FDG volumes of 125 mL or less was 0.96 (95% confidence interval, 0.91, 1.0) compared with 0.18 (95% confidence interval, 0.04, 0.85) in patients with FDG volume greater than 125 mL. Only 1 death (of leukemia) occurred in the PET-negative group (n ⫽ 66). Of the 10 patients with distant metastases and negative PET scans, all were alive and well. Patients over 45 yr with distant metastases that concentrate FDG are at the highest risk. Once distant metastases are discovered in patients with differentiated thyroid carcinoma, FDG-PET can identify high and low risk subsets. Subjects with a FDG volume greater than 125 mL have significantly reduced short term survival. (J Clin Endocrinol Metab 85: 1107–1113, 2000)

ARCINOMA arising from thyroid follicular cells is currently estimated to be present in 150,000 U.S. residents; however, it is projected that only 1,200 of these individuals will die of the disease this year (1). A variety of prognostic systems (2– 6) have arisen to identify those individuals at higher risk for death, as it would be inappropriate to treat low risk patients with therapies that may have long term side-effects. Factors such as age at diagnosis, gender, tumor factors (type, grade, and size), extrathyroidal extension, and clinical stage have been found to be useful prognostic variables (7–9). Once the disease has spread to distant locations, such as lung or bone, the aforementioned factors have not been proven to predict disease-specific survival (10). The ability of metastatic thyroid lesions to concentrate radioactive [131I]iodine (RAI) is generally held to indicate a more differentiated phenotype. Metastatic lesions that do not concentrate RAI are associated with more aggressive clinical progression (10 –13). Recent studies using [18F]fluorodeoxy-

glucose (FDG) positron emission tomography (PET) scanning have discovered that lesions that do not concentrate RAI often have elevated glucose uptake rates (14, 15). Enhanced glucose metabolism by malignant tumor tissues was first shown by Warburg in 1923 (16). With the development of FDG and large field of view PET scanners, the in vivo visualization of lesional glucose metabolism is now possible. Benign or well differentiated thyroid tumors retain FDG poorly, whereas the more malignant ones appear to have a higher uptake of FDG (14, 17). In a retrospective analysis of 125 patients followed at our medical center, we analyzed the ability of FDG-PET to identify thyroid cancer patients who may have a poor prognosis. We hypothesized that patients with metastatic lesions that did not concentrate RAI but did have high glucose uptake would have reduced survival. We discovered that survival was significantly reduced when the FDG volume was over 125 mL or the standard uptake of FDG was greater than 10 g/mL.

Received July 23, 1999. Revision received November 12, 1999. Accepted November 30, 1999. Address all correspondence and requests for reprints to: Richard J. Robbins, M.D., Endocrinology Service, Department of Medicine, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, New York 10021. * This work was supported in part by the Byrne Fund and the Laurent and Alberta Gerschel Foundation.

Patients

Subjects and Methods A total of 125 patients with differentiated thyroid cancer underwent both FDG-PET and diagnostic [131I]iodine whole body scans (DxWBS) over a 41-month period (November 1995 to March 1999). Indications for FDG-PET imaging included negative DxWBS with elevated serum thyroglobulin (Tg), high risk patients on initial presentation, and known distant metastases by other imaging modalities (plain films, magnetic

1107

1108

JCE & M • 2000 Vol 85 • No 3

WANG ET AL.

resonance imaging, computed tomography, or ultrasound). High risk was defined as age over 45 yr, less than well differentiated histology, extrathyroidal extension of tumor into adjacent neck structures, or tumor size greater than 4 cm. Not all consecutive patients fulfilling those entrance criteria had a FDG-PET study due to scheduling limitations, patient consent, and recommendation of the primary physician. Of 125 patients, 81 were female and 44 were male. The average age at diagnosis was 48.2 yr (minimum, 7 yr; maximum, 81 yr). The average age at the time of FDG-PET study was 53.3 yr (minimum, 16 yr; maximum, 84 yr). The average interval between diagnosis and FDG-PET was 63.1 months (minimum, 1.5 months; maximum, 448.5 months).

Pathological analyses Histological diagnoses were made on surgical specimens by attending pathologists at Memorial Sloan-Kettering Cancer Center (18). As a part of this study, all available primary tumors (n ⫽ 84) were reexamined for grading purposes by two of the authors (S.T. and J.R.). The grade of the lesion was judged high or low based on architectural and cytological morphology. Architectural grade was judged low if the predominant pattern was papillary and/or follicular and high if the pattern was solid, nesting (insular), or trabecular. The cytological grade was judged to be high if the nuclei were hyperchromatic and mitotically active, and/or tumor necrosis was present. Tall cell variants of papillary thyroid cancer were categorized as low grade both architecturally and cytologically.

In patients with multiple lesions, the SUV is that of the single highest lesion.

FDG volume calculation Volume estimates were obtained using a three-dimensional dosimetry software package (3D-ID) (23, 24). Contours were manually drawn to define regions of likely tumor involvement and to exclude regions of physiological FDG uptake. Within these regions, a thresholding technique (25) was used to obtain the total area of pixels whose uptake value was greater than a user-defined foreground to background contrast ratio. The sum of the area defined in each slice and the thickness of each slice gave the total tumor volume.

Statistical methods Univariate significance of all factors with respect to survival from the time of the PET scan was performed using the log-rank test for overall survival differences (26). All resulting P values are two sided. Survival curves were generated using Kaplan-Meier estimates (27). Multivariate methods were used to assess which subset of variables contributed the most information. To evaluate the level of information achieved, the score statistic obtained from the Cox proportional hazards model (28) was evaluated for each potential subset using the all subsets procedure in SAS (SAS Institute, Cary, NC).

Results

Clinical stage The clinical and pathological staging was performed as a modification of the American Joint Committee on Cancer (AJCC) system (19). The patient’s initial AJCC stage was modified only to document the presence of distant metastases at the time of entry into this study. This changed the status from M0 to M1 in 14 patients. This modification was made to provide a real-time assessment of the current clinical status. The patients had the following tumor types: 93 papillary, 18 follicular, 12 Hurthle cell (including 1 case with mixed clear cell), and 2 anaplastic carcinomas. Tumor stage was T1 in 6 cases, T2 in 20 cases, T3 in 12 cases, T4 in 69 cases, and Tx in 18 cases. The lymph node status was as follow: 26 patients were node negative (N0), 77 patients were node positive (N1 ⫽ 40; N1a ⫽ 16; N1b ⫽ 21), and 22 patients had no information concerning lymph nodes metastasis (Nx). Among the 125 patients, 58 patients had distant metastases (M1). The modified AJCC stages were as follows: 30 stage I, 21 stage II, 27 stage III, and 47 stage IV.

Treatments before FDG-PET evaluation The initial surgical procedure was total thyroidectomy in 92 patients, subtotal thyroidectomy in 13 patients, and hemithyroidectomy in 17 patients. Three patients had no surgery for the primary tumors. Before FDG-PET was performed, 101 of 125 cases had received at least 1 RAI treatment (with up to 5 treatments). The average cumulative dose of RAI before FDG-PET was 12.48 gigabecquerels (337.4 mCi), with a minimum of zero (not including the diagnostic dose) and a maximum of 64.46 gigabecquerels (1895.9 mCi).

Diagnostic RAI whole body scans DxWBS were performed as part of the thyroid dosimetry protocol of Benua and Leeper (20, 21). Five patients had dosimetry with the aid of recombinant human TSH (Genzyme Transgenics Corp., Cambridge, MA). The procedure was similar, except that the patient continued T4 treatment and received im injections of recombinant human TSH on the 2 days before the dosimetry study (22).

FDG-PET imaging The technical details of FDG-PET scanning and determination of the standard uptake value (SUV) in thyroid cancer patients were described in a previous report (17). The study was performed in 65 patients in a high TSH state (including 5 patients with rhTSH) and in 60 patients when TSH was low. We have previously reported that serum TSH levels do not alter FDG-PET outcomes in a clinically significant manner (17).

Overall survival was analyzed based on the following variables: age at diagnosis, gender, histological type and grade, presence of distant metastasis, RAI uptake, FDG uptake, maximum SUV, and total volume of FDG lesions (see Table 1). TABLE 1. Univariate analysis of time-dependent survival Variable

Age at diagnosis ⬍45 yr ⱖ45 yr Histological type Papillary Follicular Hurthle cell Anaplastic Distant metastases Absence (M0) Presence (M1) RAI uptake Positive in metastases Thyroid bed only Negative FDG-PET Positive in metastases Negative SUV (g/mL) No uptake ⱕ10 ⬎10 –20 ⬎20 PET vol (mL) None 1–125 126 –250 251–500 ⬎500

Patients

Deaths

49 76

0 14

93 18 12 2

9 2 2 1

67 58

1 13

54 22 49

8 3 3

59 66

13 1

66 32 18 9

1 4 5 4

66 40 6 8 5

1 (1.5)c 2 (5) 2 (33.3) 4 (50) 5 (100)

Logrank P valuea

0.0008 0.06

0.0001 0.34

0.0001 0.0002

0.0001b

This is not a ␹2 analysis. Due to the number of events, the statistical analysis was performed between two groups of patients: ⱕ125 and ⬎125 mL. c Percentages are in parentheses. a b

PET SCANNING AND THYROID CANCER Age and gender

Seventy-six (61%) of the patients were 45 yr or older at the time of diagnosis. By univariate analysis, those over 45 yr had a statistically significant reduced survival (P ⫽ 0.004; Fig. 1A). The gender of the patient (64% were female) did not influence survival over the 3-yr follow-up period (P ⫽ 0.16). Distant metastasis

Fifty-eight patients had lung, bone, and other distant metastases. Locoregional cervical metastases were not included as distant metastases in accordance with the AJCC staging system. All deaths due to thyroid cancer occurred in patients with distant metastases. In univariate analysis, the presence of distant metastases had a statistically significant (P ⫽ 0.0001) negative effect on survival (Fig. 1B). The single death in the group without distant metastases and in the PETnegative group was due to one patient with acute leukemia. Histological type and grade of primary tumors

Papillary thyroid cancer and its variant were diagnosed in 74% of the patients, follicular thyroid cancer in 18%, Hurthle

1109

cell carcinoma in 9%, and anaplastic carcinoma in 1% of the patients. There was a weak relationship between survival and histological type that was not significant at the 0.05 level. Furthermore, there was no significant relationship between survival and grade of the initial tumor (P ⫽ 0.60; Fig. 1C). Radioactive [131I]iodine uptake

There was no difference in survival between those with or without (including thyroid bed uptake only) RAI uptake in metastatic lesions (P ⫽ 0.34) during this short follow-up. Survival was lower in those who had both RAI and FDG uptake in distant metastases compared with those who had RAI uptake and negative FDG-PET scans (P ⫽ 0.0008; Figs. 1D and 2). FDG uptake

One death (due to acute myelocytic leukemia) occurred in the subgroup of patients with negative FDG-PET scans. All 13 patients who died of thyroid cancer had FDG-avid metastatic lesions. Univariate analysis showed that FDG uptake predicted a significantly shorter survival (P ⫽ 0.001; Fig. 3A).

FIG. 1. Kaplan-Meier plots of survival based on age (A; P ⫽ 0.0008), distant metastases (B; P ⫽ 0.0001), histological grade (C; architectural/ cytological: ⫹, high/high; ‚, high/low; E, low/low; P ⫽ 0.6), and those with distant metastases who were RAI⫺/PET⫺ (E), RAI⫹/PET⫺ (⫹), RAI⫹/PET⫹ (X), or RAI⫺/PET⫹ (‚; D; P ⫽ 0.0008).

1110

WANG ET AL.

JCE & M • 2000 Vol 85 • No 3

FIG. 2. A, Sixty-one-year-old male with poorly differentiated papillary carcinoma of thyroid. The day 7 posttherapy scan (A1) showed bilateral lung uptake. A negative FDG-PET (A2 and A3) was performed 1 week before RAI therapy. He is clinically stable. B, Sixty-six-year-old male with follicular thyroid carcinoma. The day 7 posttherapy scan (B1) showed abnormal uptake in both lungs and the left humerus. The FDG-PET (B2 and B3) performed within 1 week of RAI therapy demonstrated multiple foci of FDG-avid metastases in the lungs and mediastinum. He expired 3 yr after this PET scan despite multiple treatments with 131I, external radiation, and surgery.

Maximum SUV for FDG lesions

Among the 59 patients with positive FDG-PET scans, 27 had SUVs greater than 10 g/mL. Patients with SUV greater than 10 g/mL had a significantly shorter survival (P ⫽ 0.0002) than those with SUV of 10 g/mL or less (Fig. 3B). Removing Hurthle cell carcinoma patients from the analysis (due to their uniquely high SUV patterns) did not change the significance level (P ⫽ 0.0002).

of patients with FDG volumes of 125 mL or less was 0.96 (95% CI, 0.91, 1.0) compared with 0.18 (95% CI, 0.04, 0.85) in patients with FDG volume greater than 125 mL (Fig. 3C). In analysis of only those with distant metastases (M1), patients with FDG-avid lesions more than 125 mL had a significantly shortened survival compared to those with volumes less than 125 mL or no FDG uptake (Fig. 3D). Multivariate analyses

Volume of FDG lesions

Due to the small number of events in each volume subcategory (Table 1), patients with volumes of 125 mL or less were compared to those with volumes greater than 125 mL. The median survival, after PET scanning, with FDG volumes greater than 125 mL was 19.2 months [95% confidence interval (CI), 8.0, 34.0]. For volumes of 125 mL or less, the median was not reached. By univariate analysis, patients with FDG lesions greater than 125 mL had a significantly reduced survival (P ⫽ 0.0001). The 3-yr survival probability

Based on the univariate analyses, age, distant metastases, a positive PET scan, maximum SUV, and volume of FDG lesions were all significantly correlated with survival. To define which of these variables carried the greatest prognostic value, we performed a multivariate analysis. The score test was chosen to rank the relative strengths of all possible subsets of the five variables. The score test produced a ␹2 value of 58, corresponding to Fluorodeoxyglucose AvidDisease volume alone, whereas all other individual variables produced ␹2 values below 17. When any other variable(s)

PET SCANNING AND THYROID CANCER

1111

FIG. 3. Kaplan-Meier plots of patient survival based on abnormal retention of FDG (A; PET ⫹), volume of FDG-avid disease (B), SUV of FDG (C), and volume of FDG-avid disease in those who had distant metastases (D).

was considered with Fluorodeoxyglucose Avid-Disease volume, there was no significant rise in the value of the ␹2 test. This suggests that PET scanning in thyroid cancer patients may provide more useful information about prognosis than the presence of distant metastases. However, due to the small number of events, this multivariate analysis is considered exploratory, and the resulting model imprecise. Discussion

Prognosis in thyroid carcinoma is extremely variable. The 10-yr overall survival rates for U.S. patients with papillary, follicular, Hurthle cell, and anaplastic carcinoma are 93%, 85%, 76%, and 14%, respectively (29). Many formal prognostic systems (e.g. EORTC, AGES, and MACIS) have recognized that age at diagnosis, gender, histological features, tumor size, extrathyroidal extension, and clinical stage are key predictors of survival (6). Many individuals who die of thyroid cancer live for more than 10 yr with active disease, requiring monitoring for 20 yr or longer if they exhibit prognostic variables that influence survival (30, 31). Preliminary basic and clinical studies suggest that the most rapidly growing thyroid neoplasms have high metabolic

rates. This retrospective study monitored the health and survival of a group of 125 patients with thyroid cancer over a 41-month interval. The initial intention was to assess metabolic rate via FDG-PET scans along with other standard monitoring assays over a 10-yr period. However, the high number of deaths in certain subsets of this cohort prompted us to issue this report. We found that the total volume of FDG-avid disease correlated with prognosis and was the strongest single risk fact predicting survival in this cohort. The fact that 80% of the patients with high volume disease died during a short interval begs the question of the need for long term follow-up of this subset. Unlike larger long term studies, in which male patients have lower survival rates (5, 32), gender did not have a strong enough influence to be detected in this study (P ⫽ 0.16). The effect of age on the prognosis, however, was significant, even over this short follow-up period. All 14 deaths (including 1 not due to thyroid cancer) occurred in patients who were over 45 yr old when thyroid cancer was first diagnosed (P ⫽ 0.004). The correlation between histological type and survival is well established for thyroid carcinomas. Long term survival

1112

JCE & M • 2000 Vol 85 • No 3

WANG ET AL.

rates for papillary carcinoma are generally higher than those for follicular carcinoma, and both are significantly higher than those for anaplastic carcinoma (4, 33). As in most reports of differentiated thyroid carcinoma, the majority of our patients had papillary thyroid cancer. This histological subtype alone did not significantly influence survival in this short term follow-up study. The presence of distant metastases strongly influences survival and weighs heavily in the TNM and AMES staging systems. It is not surprising that all of the thyroid cancer deaths in our cohort occurred in those with distant metastases. However, if we stratify the 58 patients with distant metastases according to the PET findings, the clinical outcomes are different; there were no deaths in the 10 negative FDG-PET patients, whereas 13 of 48 (27%) PET-positive patients died of thyroid cancer. This was highly statistically significant (P ⫽ 0.0001). The uptake of RAI implies that some of the malignant cells retain a differentiated phenotype and can, in theory, respond to treatment with RAI. Feine et al. (14) were the first to formally propose a pattern in which an inverse relationship between RAI uptake and FDG uptake (“flip-flop”) existed in metastatic thyroid cancer lesions. However, the clinical significance of this flip-flop phenomenon has not been fully defined. Nakada et al. (13) also found that it was not the RAI uptake, but, rather, the uptake of thallium 201 that predicted the rate of growth of thyroid cancer lesions. In the present study, 88 patients had clinical evidence of local or distant metastases. Twenty-one patients had RAI uptake only (23.8%), 26 patients had FDG uptake only (29.5%), 33 patients had mixed RAI/FDG uptake (37.5%), and 8 patients had negative RAI and FDG-PET scans (9%). All of the deaths occurred in either the FDG only or the mixed RAI/FDG groups. Therefore, survival appears to be less related to the residual disease that concentrates RAI and strongly correlated with the disease that concentrates FDG. PET has established itself as an important new tool in clinical oncology. Compared to other noniodine radionuclide imaging agents, FDG-PET appears to provide the highest resolution for detecting aggressive metastatic thyroid cancer lesions (34 –36). Although the utility of FDG-PET for detecting thyroid cancer lesions in patients who have negative DxWBS and elevated serum Tg has been established (17), its potential as a prognostic marker has yet to be determined. A recent study demonstrated that primary tumor SUV was the single best predictor for survival in nonsmall cell lung carcinoma (37). In addition to showing reduced survival in those with FDG-avid disease, we also found a significant correlation with the maximum SUV, suggesting that tumors with the highest metabolic activity might be those with the most rapid growth potential. Finally, we explored a new approach by measuring the three-dimensional FDG volume (23, 25). We discovered that survival was directly related to the total volume of FDG disease; the higher the volume, the shorter the survival. Multivariate analyses revealed that high volume (⬎125 mL) of the FDG-avid diseases provided stronger prognostic information than did age, gender, initial histological type or grade, presence of RAI uptake, or modified AJCC stage. In fact, the strongest single predictor of short term survival was

the volume of FDG-avid tumor. Eighty percent of those with FDG-PET volumes over 125 mL had expired by 41 months. We do not have sufficient data to estimate the rate of progression of FDG volume; however, strong consideration should be given to periodic FDG-PET scanning in patients with metastatic thyroid carcinoma who have FDG-positive lesions. Unfortunately, there is no established treatment that will improve survival in patients with metastatic FDG-PETpositive disease. Our preliminary data suggest that high dose RAI is not able to eradicate FDG-avid metastases (38). This is in agreement with a recent report on the inability of radioiodine to eradicate thyroid cancer lesions that take up high levels of thallium 201 (13). In summary, we identified a group of patients with metastatic thyroid cancer at very high risk for dying within a 41-month interval. These patients had distant metastases that concentrated FDG regardless of whether they also concentrated RAI. The presence of FDG uptake was not as strong a predictor of survival as was the FDG-PET volume of disease. Therapeutic research protocols for such high risk patients need to be designed, possibly using FDG-PET assessments as quantitative outcome measures. Acknowledgments We thank the Byrne family for providing support via the Byrne Fund Cancer Research Program, and the Laurent and Alberta Gerschel Foundation for their foresight in supporting the PET Center at Memorial Hospital. The completion of this work was supported by the enthusiasm and dedication of our long term friends, Joan Shey and Adele Platzer, and their Light of Life Foundation, which is dedicated to improving the quality of life for thyroid cancer patients. We also thank Cherryl MurrayMarone for valuable editorial assistance.

References 1. Landis S, Murray T, Bolden S, Wingo P. 1999 Cancer statistics, 1999. CA Cancer J Clinicians. 49:8 –31. 2. Tennvall J, Bjorkland A, Moller T, et al. 1986 Is the EORTC prognostic index of thyroid cancer valid in differentiated thyroid carcinoma: retrospective multivariate analysis of differentiated thyroid carcinoma with long follow-up. Cancer. 57:1405–1414. 3. Cady B, Rossi R. 1988 An expanded view of risk-group definition in differentiated thyroid carcinoma. Surgery. 104:947–953. 4. Hay I, Bergstralh E, Goellner J, Ebersold J, Grant C. 1993 Predicting outcome in papillary thyroid carcinoma: development of a reliable prognostic scoring system in a cohort of 1779 patients surgically treated at one institution during 1940 through 1989. Surgery. 114:1050 –1058. 5. Akslen L, Myking A, Salvesen H, et al. 1993 Prognostic importance of various clinicopathological features in papillary thyroid carcinoma. Eur J Cancer. 29:44 –54. 6. Sherman S. 1999 Toward a standard clinocopathologic staging approach for differentiated thyroid carcinoma. Semin Surg Oncol. 16:12–15. 7. Ain K. 1995 Papillary thyroid carcinoma. Endocrinol Metab Clinics North Am. 24:711–760. 8. Hay I. 1989 Prognostic factors in thyroid carcinoma. Thyroid Today. 12:1–9. 9. Grebe SK, Hay I. 1995 Follicular thyroid cancer. Endocrinol Metab Clinics North Am. 24:761– 801. 10. Schlumberger M. 1998 Papillary and follicular thyroid carcinoma. N Engl J Med. 338:297–306. 11. Hoie J, Stenwig A, Kullmann G, Lindegaard M. 1988 Distant metastases in papillary thyroid cancer. A review of 91 patients. Cancer. 61:1– 6. 12. Segal R. 1992 Radioactive iodine treatment of thyroid carcinoma. In: Cobin R, Sirota D, ed. Malignant tumors of the thyroid. New York: Springer-Verlag; 100 –111. 13. Nakada K, Katoh C, Kanegae K, et al. 1998 Thallium-201 scintigraphy to predict therapeutic outcome of iodine-131 therapy of metastatic thyroid carcinoma. J Nucl Med. 39:807– 810. 14. Feine U, Lietzenmayer R, Hanke J, Held J, Wohrle H. 1996 Fluorine-18-FDG and iodine-131 uptake in thyroid cancer. J Nucl Med. 37:1468 –1472. 15. Dietlein M, Scheidhauer K, Voth E, Theissen P, Schicha H. 1997 Fluorine-18 fluorodeoxyglucose positron emission tomography and iodine-131 whole

PET SCANNING AND THYROID CANCER

16. 17.

18. 19. 20. 21. 22.

23. 24. 25. 26.

body scintigraphy in the follow-up of differentiated thyroid cancer. Eur J Nucl Med. 24:1342–1348. Warburg O. 1923 Versuche und uberledbeudem carcinomgewebe (methoden). Biochem Z. 142:317–333. Wang W, Macapinlac H, Finn R, et al. 1999 PET scanning with 18F-fluoro2-deoxyglucose can localize residual differentiated thyroid cancer in patients with negative 131Iodine whole body scans. J Clin Endocrinol Metab. 84:2291– 2302 . Rosai J, Carcangiu M, DeLellis R. 1992 Tumors of the thyroid gland. In: Rosai J, ed. Atlas of tumor pathology. Washington DC: Armed Forces Institute of Pathology; 21–206. AJCC Clinical Staging Manual. 1997 American Joint Commission on Cancer, 5th Ed, vol 2. Philadelphia, New York: Lippincott-Raven. Benua R, Cicale N, Sonenberg M, Rawson R. 1962 The relation of radioiodine dosimetry to results and complications in the treatment of metastatic thyroid cancer. Am J Roentgenol. 87:171–179. Leeper R. 1973 The effect of 131I therapy on survival of patients with metastatic papillary or follicular thyroid carcinoma. J Clin Endocrinol Metab. 36:1143–1152. Ladenson P, Braverman L, Mazzaferri E, et al. 1997 Comparison of administration of recombinant human thyrotropin with withdrawal of thyroid hormone for radioactive iodine scanning in patients with thyroid carcinoma. N Engl J Med. 337:888 – 896. Kolbert KS, Sgouros G, Scott AM, et al. 1997 Implementation and evaluation of patient-specific three-dimensional internal dosimetry. J Nucl Med. 38:301–308. Sgouros G, Chiu S, Pentlow KS, et al. 1993 Three-dimensional dosimetry for radioimmunotherapy treatment planning. J Nucl Med. 34:1595–1601. Erdi Y, Mawlawi O, Larson S, et al. 1997 Segmentation of lung lesion volume by adaptive positron emission tomography image thresholding. Cancer. 80(Suppl):2505–2509. Heller G, Venkatraman ES. 1996 Resampling procedures to compare two survival distributions in the presence of right-censored data. Biometrics. 52:1204 –1213.

1113

27. Kaplan E, Meier P. 1958 Nonparametric estimation from incomplete observations. J Am Stat Assoc. 53:457– 481. 28. Cox DR, Oakes D. 1974 Analysis of survival data. London: Chapman and Hall. 29. Hundahl S, Fleming I, Fremgen A, Menck H. 1998 A national cancer data base report on 53,856 cases of thyroid carcinoma treated in the US, 1985–1995. Cancer. 83:2638 –2648. 30. Mazzaferri E., Jiang S. 1994 Long-term impact of initial surgical and medical therapy on papillary and follicular thyroid cancer. Am J Med. 97:418 – 428. 31. Schlumberger M, Challeton C, De Vathaire F, et al. 1996 Radioactive iodine treatment and external radiotherapy for lung and bone metastases from thyroid carcinoma. J Nucl Med. 37:598 – 605. 32. Schindler A, van Melle G, Evequoz B, et al. 1991 Prognostic factors in papillary carcinoma of the thyroid. Cancer. 68:324 –330. 33. Tubiana M, Schlumberger M, Rougier P, et al. 1985 Long term results and prognostic in patients with differentiated thyroid carcinoma. Cancer. 55:794 – 804. 34. Uematsu H, Sadato N, Ohtsubo T, et al. 1998 Fluorine-18-fluorodeoxyglucose PET vs. thallium-201 scintigraphy evaluation of thyroid tumors. J Nucl Med. 39:453– 459. 35. Grunwald F, Menzel C, Bender H, et al. 1997 Comparison of 18FDG-PET with 131iodine and 99 mTc-sestamibi scintigraphy in differentiated thyroid cancer. Thyroid. 7:327–335. 36. Lind P, Gallowitsch H, Unterweger O, et al. 1998 FDG PET in the follow-up of thyroid cancer: comparison with Tc-99 m tetrofosmin and 131-I whole body scintigraphy. Thyroid. 8:1221–1222. 37. Ahuja V, Coleman R, Herndon J, Patz E. 1998 The prognostic significance of fluorodeoxyglucose positron emission tomography imaging for patients with non-small cell lung carcinoma. Cancer. 83:918 –924. 38. Wang W, Larson S, Turlakow A, Yeung H, Macapinlac H, Kalagian H, Robbins R. 1999 Resistance of FDG-PET positive metastatic thyroid carcinoma lesions to treatment with radioactive iodine [Abstract 191]. Thyroid. 9(Suppl 1):32.